Tech Guides - Artificial Intelligence

The short answer is yes, and no. The long answer is, well, read on to find out. Several have been asking the question, including myself, wondering whether Artificial Intelligence is just another passing fad like maybe the Google Glass or nano technology. The hype for AI began over the past few years, although if you actually look back at the 60’s it seems to have started way back then. In the early 90s and all the way down to the early 2000’s, a lot of media and television shows were talking about AI quite a bit. Going 25 centuries even further back, Aristotle speaks of not just thinking machines but goes on to talk of autonomous ones in his book, Politics: for if every instrument, at command, or from a preconception of its master's will, could accomplish its work (as the story goes of the statues of Daedalus; or what the poet tells us of the tripods of Vulcan, "that they moved of their own accord into the assembly of the gods "), the shuttle would then weave, and the lyre play of itself; nor would the architect want servants, or the [1254a] master slaves. Aristotle, Politics: A treatise on Government, Book 1, Chapter 4 This imagery of AI has managed to sink into our subconscious minds over the centuries propelling creative work, academic research and industrial revolutions toward that goal. The thought of giving machines a mind of their own, existed quite long ago, but recent advancements in technology have made it much clearer and realistic. The Rise of the Machines The year is 2018. The 4th Industrial Revolution is happening and intelligent automation has taken over. This is the point where I say no, AI is not overhyped. General Electric, for example, is a billion dollar manufacturing company that has already invested in AI. GE Digital has AI systems running through several automated systems. They even have their own IIoT platform called Predix. Similarly, in the field of healthcare, the implementation of AI is growing in leaps and bounds. The Google Deepmind project is able to process millions of medical records within minutes. Although this kind of research is in its early phase, Google is working closely with the Moorfields Eye Hospital NHS Foundation Trust to implement AI and improve eye treatment. AI startups focused on healthcare and other allied areas such as genetic engineering are some of the highly invested and venture capital supported ones in recent times. Computer Vision or image recognition is one field where AI has really proven its power. Analysing datasets like iris has never been easier, paving way for more advanced use cases like automated quality checks in manufacturing units. Another interesting field is Healthcare, where AI has helped sift through tonnes of data, helping doctors diagnose illnesses quicker, manufacture more effective and responsive drugs, and in patient monitoring. The list is endless, clearly showing that AI has made its mark in several industries. Back (up) to the Future Now, if you talk about the commercial implementations of AI, they’re still quite far fetched at the moment. Take the same Computer Vision application for example. Its implementation will be a huge breakthrough in autonomous vehicles. But if researchers have managed to obtain an 80% accuracy for object recognition on roads, the battle is not close to being won! Even if they do improve, do you think driverless vehicles are ready to drive in the snow, through the rain or even storms? I remember a few years ago, Business Process Outsourcing was one industry, at least in India, that was quite fearful of the entry of AI and autonomous systems that might take over their jobs. Machines are only capable of performing 60-70% of the BPO processes in Insurance, and with changing customer requirements and simultaneously falling patience levels, these numbers are terrible! It looks like the end of Moore’s law is here, for AI I mean. Well, you can’t really expect AI to have the same exponential growth that computers did, decades ago. There are a lot of unmet expectations in several fields, which has a considerable number of people thinking that AI isn’t going to solve their problems now, and they’re right. It is probably going to take a few more years to mature, making it a thing of the future, not of the present. Is AI overhyped now? Yeah, maybe? What I think Someone once said, hype is a double-edged sword. If it’s not enough, innovation may become obscure and if it’s too much, expectations will become unreasonable. It’s true that AI has several beneficial use cases, but what about fairness of such systems? Will machines continue to think the way they’re supposed to or will they start finding their own missions that don’t involve benefits to the human race? At the same time, there’s also a question of security and data privacy. GDPR will come into effect in a few days, but what about the prevailing issues of internet security? I had an interesting discussion with a colleague yesterday. We were talking about what the impact of AI could be for us as end-customers, in a developing and young country like India. Do we really need to fear losing our jobs, will we be able to reap the benefits of AI directly or would it be an indirect impact? The answer is, probably yes, but not so soon. If we drew up the hierarchy of needs pyramid for AI, it would look something like the above. For each field to fully leverage AI, it’s going to involve several stages like collecting data, storing it effectively, exploring it, then aggregating it, optimising it with the help of algorithms and then finally achieving AI. That’s bound to take a LOT of time! Honestly speaking, a country like India lacks as much implementation of AI in several fields. The major customers of AI, apart from some industrial giants, will obviously be the government. Although, that is sure to take at least a decade or so, keeping in mind the several aspects to be accomplished first. In the meantime, buddying AI developers and engineers are scurrying to skill themselves up in the race to be in the cream of the crowd! Similarly, what about the rest of the world? Well, I can’t speak for everyone, but if you ask me, AI is a really promising technology and I think we need to give it some time; allow the industries and organisations investing in it to take enough time to let it evolve and ultimately benefit us customers, one way or another. You can now make music with AI thanks to Magenta.js Splunk leverages AI in its monitoring tools    

Chatbots are remarkably changing the way customer service is provided in a variety of industries. For every organization, customer satisfaction plays a very important role, thus they expect business to be reachable any time and respond to their queries 24*7. With growing artificial intelligence advances in smart devices and IoT, chatbots are becoming a necessity for communicating with customers in real time. There are many existing vendors such as Google, Microsoft, Amazon, and IBM with the required models and services to build conversational interfaces for the applications and devices. But the chatbot industry is evolving and even minor improvements in the UI, or the algorithms that work behind the scenes or the data they use to get trained, can mean a major win. With complete backing by the Facebook team, we can expect Wit.ai creating new simplified ways to ease speech recognition and voice interface for developers.  Wit.ai has an excellent support for NLP making it one of the popular bot frameworks in the market. The key to chatbot success is to pursue continuous learning that enables them to leverage relevant data in order to connect with clearly defined customers, this what makes Wit.ai extra special. What is Wit.ai? Wit.ai is an open and extensible NLP engine for developers, acquired by Facebook, which allows you to build conversational applications and devices that you can talk or text to. It provides an easy interface and quick learning APIs to understand human communication from every interaction and helps to parse the complex message (which can be either voice or text) into structured data. It also helps you with predicting the forthcoming set of events based on the learning from the gathered data. Why Wit.ai It is one of the most powerful APIs used to understand natural language It is a free SaaS platform that provides services for developers to build a chatbot for their app or device. It has story support thus allowing you to visualize the user experience. A new built-in support NLP integration with the Page inbox allows the page admins to create a Wit app with ease. Further by using the anonymized samples from past messages, the bot provides automate responses to the most common requests asked. You can create efficient and powerful text or voice based conversational bots that humans can chat with. In addition to business bots, these APIs can be used to build hands-free voice interfaces for mobile phones, wearable devices, home automation products and more. It can be used in platforms that learn new commands semantically to those input by the developer. It provides a developer GUI which includes a visual representation of the conversation flows, business logic invocations, context variables, jumps, and branching logic. Programming language and integration support - Node.js client, Python client, Ruby client, and HTTP API. Challenges in Wit.ai Wit.ai doesn’t support third-party integration tools. Wit.ai has no required slot/parameter feature. Thus you will have to invoke business logic every time there is an interaction with the user in order to gather any missing information not spoken by the user. Training the engine can take some time based on the task performed. When the number of stories increases, Wit engine becomes slower. However, existing Wit.ai adoption looks very promising, with more than 160,000 members in the community contributing on GitHub. In order to have a  complete coverage of tutorials, documentation and client support APIs you can visit the Github page to see a list of repositories. My friend, the robot: Artificial Intelligence needs Emotional Intelligence Snips open sources Snips NLU, its Natural Language Understanding engine What can Google Duplex do for businesses?  

After deep learning, reinforcement Learning (RL), the hottest branch of Artificial Intelligence that is finding speedy adoption in tech-driven companies. Simply put, reinforcement learning is all about algorithms tracking previous actions or behaviour and providing optimized decisions using trial-and-error principle. Read How Reinforcement Learning works to know more. It might sound theoretical but gigantic firms like Google and Uber have tested out this exceptional mechanism and have been highly successful in cutting edge applied robotics fields such as self driving vehicles. Other top giants including Amazon, Facebook and Microsoft have centralized their innovations around deep reinforcement learning across Automotive, Supply Chain, Networking, Finance and Robotics. With such humongous achievement, reinforcement learning libraries has caught the Artificial Intelligence developer communities’ eye and have gained prime interest for training agents and reinforcing the behavior of the trained agents. In fact, researchers believe in the tremendous potential of reinforcement learning to address unsolved real world challenges like material discovery, space exploration, drug discovery etc and build much smarter artificial intelligence solutions. In this article, we will have a look at the most promising open source tools and libraries to start building your reinforcement learning projects on. OpenAI Gym OpenAI Gym, the most popular environment for developing and comparing reinforcement learning models, is completely compatible with high computational libraries like TensorFlow. The Python based rich AI simulation environment offers support for training agents on classic games like Atari as well as for other branches of science like robotics and physics such as Gazebo simulator and MuJoCo simulator. The Gym environment also offers APIs which facilitate feeding observations along with rewards back to agents. OpenAI has also recently released a new platform, Gym Retro made up of 58 varied and specific scenarios from Sonic the Hedgehog, Sonic the Hedgehog 2, and Sonic 3 games. Reinforcement learning enthusiasts and AI game developers can register for this competition. Read: How to build a cartpole game using OpenAI Gym TensorFlow This is an another well-known open-source library by Google followed by more than 95,000 developers everyday in areas of natural language processing, intelligent chatbots, robotics, and more. The TensorFlow community has developed an extended version called TensorLayer providing popular RL modules that can be easily customized and assembled for tackling real-world machine learning challenges. The TensorFlow community allows for the framework development in most popular languages such as Python, C, Java, JavaScript and Go. Google & its TensorFlow team are in the process of coming up with a Swift-compatible version to enable machine learning  on Apple environment. Read How to implement Reinforcement Learning with TensorFlow Keras Keras presents simplicity in implementing neural networks with just a few lines of codes with faster execution. It provides senior developers and principal scientists with a high-level interface to high tensor computation framework, TensorFlow and centralizes on the model architecture. So, if you have any existing RL models written in TensorFlow, just pick the Keras framework and you can transfer the learning to the related machine learning problem. DeepMind Lab DeepMind Lab is a Google 3D platform with customization for agent-based AI research. It is utilized to understand how self-sufficient artificial agents learn complicated tasks in large, partially observed environments. With the victory of its AlphaGo program against go players, in early 2016, DeepMind captured the public’s attention. With its three hubs spread across London, Canada and France, the DeepMind team is focussing on core AI fundamentals which includes building a single AI system backed by state-of-the-art methods and distributional reinforcement learning. To know more about how DeepMind Lab works, read How Google’s DeepMind is creating images with artificial intelligence. Pytorch Pytorch, open sourced by Facebook, is another well-known deep learning library adopted by many reinforcement learning researchers. It was recent preferred almost unanimously by top 10 finishers in Kaggle competition. With dynamic neural networks and strong GPU acceleration, Rl practitioners use it extensively to conduct experiments on implementing policy-based agent and to create new adventures. One crazy research project is Playing GridWorld, where Pytorch unchained its capabilities with renowned RL algorithms like policy gradient and simplified Actor-Critic method. Summing It Up There you have it, the top tools and libraries for reinforcement learning. The list doesn't end here, as there is a lot of work happening in developing platforms and libraries for scaling reinforcement learning. Frameworks like RL4J, RLlib are already in development and very soon would be full-fledged available for developers to simulate their models in their preferred coding language.

When talking about the capabilities of AI-driven digital assistants, the most talked about issue is their inability to converse in a way a real human does. The robotic tone of the virtual assistants has been limiting them from imitating real humans for a long time. And it’s not just the flat monotone. It’s about understanding the nuances of the language, pitches, intonations, sarcasm, and a lot more. Now, what if there emerges a technology that is capable of sounding and behaving almost human? Well, look no further, Google Duplex is here to dominate the world of digital assistants. Google introduced the new Duplex at Google I/O 2018, their annual developer conference, last week. But, what exactly is it? Google Duplex is a newly added feature to the famed Google assistant. Adding to the capabilities of Google assistant, it is also able to make phone calls for the users, and imitate human natural conversation almost perfectly to get the day-to-day tasks ( such as booking table reservations, hair salon appointments, etc. ) done in an easy manner. It includes pause-fillers and phrases such as “um”, “uh-huh “, and “erm” to make the conversation sound as natural as possible. Don’t believe me? Check out the audio yourself! [audio mp3="https://hub.packtpub.com/wp-content/uploads/2018/05/Google-Duplex-hair-salon.mp3"][/audio]  Google Duplex booking appointments at a hair salon [audio mp3="https://hub.packtpub.com/wp-content/uploads/2018/05/Google-Duplex-table-reservation.mp3"][/audio]  Google Duplex making table reservations at a restaurant The demo call recording video of the assistant and the business employee, presented by Sundar Pichai, Google’s CEO, during the opening keynote, befuddled the entire world about who’s the assistant and who’s the human, making it go noticeably viral. A lot of questions are buzzing around whether Google Duplex just passed the Turing Test. The Turing Test assesses a machine’s ability to present intelligence closer or equivalent to that of a human being. Did the new human sounding robot assistant pass the Turing test yet? No, but it’s certainly the voice AI that has come closest to passing it. Now how does Google Duplex work? It’s quite simple. Google Duplex finds out the information ( you need ) that isn’t out there on the internet by making a direct phone call. For instance, a restaurant has shifted location and the new address is nowhere to be found online. Google Duplex will call the restaurant and check on their new address for you. The system comes with a self-monitoring capability, helping it recognize complex tasks that it cannot accomplish on its own. Such cases are signaled to a human operator, who then takes care of the task. To get a bit technical, Google Duplex makes use of Recurrent Neural Networks ( RNNs ) which are created using TensorFlow extended ( TFX ), a machine learning platform. Duplex’s RNNs are trained using the data anonymization technique on phone conversation data. Data anonymization helps with protecting the identity of a company or an individual by removing the data sets related to them. The output of Google’s Automatic speech recognition technology, conversation history and different parameters of the conversation are used by the network. The model also makes use of hyperparameter optimization from TFX which further enhances the model. But, how does it sound natural? Google uses concatenative text to speech ( TTS ) along with synthesis TTS engine ( using Tacotron and WaveNet ) to control the intonation depending on different circumstances. Concatenative TTS is a technique that converts normal text into speech by concatenating or linking together the recorded speech pieces. Synthesis TTS engine helps developers modify the speech rate, volume, and pitch of the synthesized output. Including speech disfluencies ( “hmm”s, “erm”s, and “uh”s ) makes the Duplex sound more human. These speech disfluencies are added when very different sound units are combined in the concatenative TTS or adding synthetic waits. This allows the system to signal in a natural way that it is still processing ( equivalent to what humans do when trying to sort out their thoughts ). Also, the delay or latency should match people’s expectations. Duplex is capable of figuring out when to give slow or fast responses using low-confidence models or faster approximations. Google also found out that including more latency helps with making the conversation sound more natural. Some potential applications of Google Duplex for businesses Now that we’ve covered the what and how of this new technology, let’s look at five potential applications of Google Duplex in the immediate future. Customer Service Basic forms of AI using natural language processing ( NLP ), such as chatbots and the existing voice assistants such as Siri and Alexa are already in use within the customer care industry. Google Duplex paves the way for an even more interactive form of engaging customers and gaining information, given its spectacular human sounding capability. According to Gartner, “By 2018, 30% of our interactions with technology will be through "conversations" with smart machines”. With Google Duplex, being the latest smart machine introduced to the world, the basic operations of the customer service industry will become easier, more manageable and efficient. From providing quick solutions to the initial customer support problems and delivering internal services to the employees, Google Duplex perfectly fills the bill. And it will only get better with further advances in NLP. So far chatbots and digital assistants have been miserable at handling irate customers. I can imagine Google Duplex in John Legend’s smooth voice calming down an angry customer or even making successful sales pitches to potential leads with all its charm and suave! Of course, Duplex must undergo the right customer management training with a massive amount of quality data on what good and bad handling look like before it is ready for such a challenge. Other areas of customer service where Google Duplex can play a major role is in IT support. Instead of connecting with the human operator, the user will first get connected to Google Duplex. Thus, making the entire experience friendly and personalized from the user perspective and saving major costs for organizations. HR Department Google Duplex can also extend a helping hand in the HR department. The preliminary rounds of talent acquisition where hiring executives make phone calls to their respective candidates could be handled by Google Duplex provided it gets the right training. Making note of the basic qualifications, candidate details, and scheduling interviews are all the functions that Google Duplex should be able to do effectively. The Google Assistant can collect the information and then further rounds can be conducted by the human HR personnel. This could greatly cut down on the time expended by HR executives on the first few rounds of shortlisting. This means they are free to focus their time on other strategically important areas of hiring. Personal assistants and productivity As presented at Google I/O 2018, Google Duplex is capable of booking appointments at hair salons, booking table reservations and finding out holiday hours over the phone. It is not a stretch to therefore assume that it can also order takeaway food over a phone call, check with the delivery man regarding the order, cancel appointments, make business inquiries, etc. Apart from that, it’s a great aid for people with hearing loss issues as well as people who do not speak the local language by allowing them to carry out tasks on phone. Healthcare Industry There is already enough talk surrounding the use of Alexa, Siri, and other voice assistants in healthcare. Google Duplex is another new addition to the family. With its natural way of conversing, Duplex can: Let patients know their wait time for emergency rooms. Check with the hospital regarding their health appointments. Order the necessary equipment for hospital use. Another allied area is elder care. Google Duplex could help reduce ailments related to loneliness by engaging with the users at a more human level. It could also assist with preventive care and in the management of lifestyle diseases such as diabetes by ensuring patients continue their med intake, keep their appointments, provide emergency first aid help, call 911 etc. Real Estate Industry Duplex enabled Google Assistants will help make realtors’ task easy. Duplex can help call potential sellers and buyers, thereby, making it easy for realtors to select the respective customers. The conversation between Google Duplex ( helping a realtor ) and a customer wanting to buy a house can look something like this: Google Duplex: Hi! I heard you are house hunting. Are you looking to buy or sell a property? Customer: Hey, I’m looking to buy a home in the Washington area. Google Duplex: That’s great! What part of Washington are you looking in for? Customer:  I’m looking for a house in Seattle. 3 bedrooms and 3 baths would be fine. Google Duplex: Sure, umm, may I know your budget? Customer: Somewhere between $749,000 to $850,000, is that fine? Google Duplex: Ahh okay sure, I’ve made a note and I’ll call you once I find the right matches. Customer: Yeah, sure. Google Duplex: okay, thanks. Customer: Thanks, Bye! Google Duplex then makes a note of the details on the realtor’s phone, thereby, narrowing down the efforts made by realtors on cold calling the potential sellers to a great extent. At the same time, the broker will also receive an email with the consumer’s details and contact information for a follow-up. Every rose has its thorns. What’s Duplex’s thorny issue? With all the good hype surrounding Google Duplex, there have been some controversies regarding the ethicality of Google Duplex. Some people have questions and mixed reactions about Google Duplex fooling people of one’s identity as the voice of the Duplex differs significantly from that of a robot. A lot of talk surrounding this issue is trending on several twitter threads. It has hushed away these questions by saying how ‘transparency in technology’ is important and they are ‘designing this feature with disclosure built-in’ which will help in identifying the system. Google also mentioned how any feedback that people have regarding their new product. Google successfully managed to awe people across the globe with their new and innovative Google Duplex. But there is a still a long way to go even though Google has already taken a step ahead in an effort to better the human relationships with the machines. If you enjoyed reading this article and want to know more, check out the official Google Duplex blog post. Google’s Android Things, developer preview 8: First look Google News’ AI revolution strikes balance between personalization and the bigger picture Android P new features: artificial intelligence, digital wellbeing, and simplicity  

‘Can Recurrent neural networks warp time?’ is authored by Corentin Tallec and Yann Ollivier to be presented at ICLR 2018. This paper explains that plain RNNs cannot account for warpings, leaky RNNs can account for uniform time scalings but not irregular warpings, and gated RNNs can adapt to irregular warpings. Gating mechanism of LSTMS (and GRUs) to time invariance / warping What problem is the paper trying to solve? In this paper, that authors prove that learnable gates in a recurrent model formally provide quasi-invariance to general time transformations in the input data. Further, the authors try to recover part of the LSTM architecture from a simple axiomatic approach. This leads to a new way of initializing gate biases in LSTMs and GRUs. Experimentally, this new chrono initialization is shown to greatly improve learning of long term dependencies, with minimal implementation effort. Paper summary The authors have derived the self loop feedback gating mechanism of recurrent networks from first principles via a postulate of invariance to time warpings. Gated connections appear to regulate the local time constants in recurrent models. With this in mind, the chrono initialization, a principled way of initializing gate biases in LSTMs, has been introduced. Experimentally, chrono initialization is shown to bring notable benefits when facing long term dependencies. Key takeaways In this paper, the authors show that postulating invariance to time transformations in the data (taking invariance to time warping as an axiom) necessarily leads to a gate-like mechanism in recurrent models. The paper provides precise prescriptions on how to initialize gate biases depending on the range of time dependencies to be captured. The empirical benefits of the new initialization on both synthetic and real world data have been tested. The authors also observed a substantial improvement with long-term dependencies, and slight gains or no change when short-term dependencies dominate. Reviewer comments summary Overall Score: 25/30 Average Score: 8 According to a reviewer, the core insight of the paper is the link between recurrent network design and its effect on how the network reacts to time transformations. This insight is simple, elegant and valuable, as per the reviewer. A minor complaint highlighted is that there are an unnecessarily large number of paragraph breaks, which make reading slightly jarring. Recurrent neural networks and the LSTM architecture Build a generative chatbot using recurrent neural networks (LSTM RNNs) How to recognize Patterns with Neural Networks in Java  

Training a machine learning model to give accurate results requires crunching huge amounts of labelled data in it. Data being naturally unlabelled, need ‘experts’ who can scan through the data and tag them with correct labels. To perform topic-specific data labelling, for example, classifying diseases based on their type, would definitely require a doctor or someone with a medical background to label the data. Getting such topic-specific experts to label data can get difficult and quite expensive. Also, doing this for many machine learning projects is impractical. Active learning can help here. What is Active Learning Active learning is a type of semi-supervised machine learning, which aids in reducing the amount of labeled data required to train a model. In active learning, the model focuses only on data that the model is confused about and requests the experts to label them. The model later trains a bit more on the small amount of labeled data, and repeats the same for such confusing data labeling. Active learning, in short, prioritizes confusing samples that need labeling. This enables models to learn faster, and allows experts to skip labeling data that is not a priority, and to provide the model with the most useful information on the confused samples. This in turn can fetch great machine learning models, as active learning can reduce the number of labels required to collect from experts. Types of Active learning An active learning environment includes a learner (the model being trained), huge amount of raw and unlabelled data, and the expert (the person/system labelling the data). The role of the learner is to choose which instances or examples should be labelled. The learner’s goal is to reduce the number of labeled examples needed for an ML model to learn. On the other hand, the expert on receiving the data to be labelled, analyzes the data to determine appropriate labels for it. There are three types of Active learning scenarios. Query Synthesis - In such a scenario, the learner constructs examples, which are further sent to the expert for labeling. Stream-based active learning - Here, from the stream of unlabelled data, the learner decides the instances to be labelled or choose to discard them. Pool-based active learning - This is the most common scenario in active learning. Here, the learner chooses only the most informative or best instances and forwards them to the expert for labelling. Some Real-life applications of Active learning Natural Language Processing (NLP): Most of the NLP applications require a lot of labelled data such as POS (Parts-of-speech) tagging, NER (Named Entity Recognition), and so on. Also, there is a huge cost incurred in labelling this data. Thus, using active learning can reduce the amount of labelled data required to label. Scene understanding in self-driving cars: Active learning can also be used in detecting objects, such as pedestrians from a video camera mounted on a moving car,a key area to ensure safety in autonomous vehicles. This can result in high levels of detection accuracy in complex and variable backgrounds. Drug designing: Drugs are biological or chemical compounds that interact with specific ‘targets’ in the body (usually proteins, RNA or DNA) with an aim to modify their activity. The goal of drug designing is to find which compounds bind to a particular target. The data comes from large collections of compounds, vendor catalogs, corporate collections, and combinatorial chemistry. With active learning, the learner can find out the compounds that are active (binds to target) or inactive. Active learning is still being researched using different deep learning algorithms such as CNNs and LSTMs, which act as learners in order to improve their efficiency. Also, GANs (Generative Adversarial Networks) are being implemented in the active learning framework. There are also some research papers that try to learn active learning strategies using meta-learning. Why is Python so good for AI and Machine Learning? 5 Python Experts Explain AWS Greengrass brings machine learning to the edge Unity Machine Learning Agents: Transforming Games with Artificial Intelligence

Woah, woah, woah! Wait a minute! First there was that game SpecOps that I usually sucked at, then there came ITOps and DevOps that took the world by storm, now there’s another something-Ops?? Well, believe it or not, there is, and they’re calling it AIOps. What does AIOps stand for? AIOps basically means Artificial Intelligence for IT Operations. It means IT operations are enhanced by using analytics and machine learning to analyze the data that’s collected from various IT operations tools and devices. This helps in spotting and reacting to issues in real time. Coined by Gartner, the term has grown in popularity over the past year. Gartner believes that AIOps will be a major transformation for ITOps professionals mainly due to the fact that traditional IT operations cannot cope with the modern digital transformation. Why is AIOps important? With the massive and rapid shift towards cloud adoption, automation and continuous improvement, AIOps is here to take care of the new entrants into the digital ecosystem - Machine agents, artificial intelligence, IoT devices, etc. These new entrants are impossible to service and maintain by humans and with billions of devices connected together, the only way forward is to employ algorithms that tackle known problems. Some of the solutions it provides are maintaining high availability and monitoring performance, event correlation and analysis, automation and IT service management. How does AIOps work? As depicted in Gartner’s diagram, there are two primary components to AIOps. Big Data Machine Learning Data is gathered from the enterprise. You then implement a comprehensive analytics and machine learning strategy alongside the combined IT data (monitoring data + job logs + tickets + incident logs). The processed data yields continuous insights, continuous improvements and fixes. It bridges three different IT disciplines to accomplish its goals: Service management Performance management, and Automation To put it simply, it is a strategic focus. It argues for a new approach in a world where big data and machine learning have changed everything. How to move from ITOps to AIOps Machine Learning Most of AIOps will involve supervised learning and professionals will need a good understanding of the underlying algorithms. Now don’t get me wrong, they don’t need to be full blown data scientists to build the system, but just having sufficient knowledge to be able to train the system to pick up anomalies. Auditing these systems to ensure they’re performing the tasks as per the initial vision is necessary and this will go hand in hand with scripting them. Understanding modern application technologies With the rise of Agile software development and other modern methodologies, AIOps professionals are expected to know all about microservices, APIs, CI/CD, containers, etc. With the giant leaps that cloud development is taking, it is expected to gain visibility into cloud deployments, with an emphasis on cost and performance. Security Security is critical, for example, it’s important for personnel to understand how to engage a denial of service attack or maybe a ransomware attack, like the ones we’ve seen in the recent past. Training machines to detect/predict such events is pertinent to AIOps. The key tools in AIOps There are a wide variety of AIOps platforms available in the market that bring AI and Intelligence to IT Operations. One of the most noteworthy ones is Splunk, which has recently incorporated AI for intelligence driven operations. Another one is the Moogsoft AIOps platform, that is quite similar to Splunk. BMC also has entered the fray, launching TrueSight 11, their AIOps platform that promises to address use cases to improve performance and capacity management, the service desk, and application development disciplines. Gartner has a handy list of top platforms. If you’re planning the transition from ITOps, do check out the list. Companies like Frankfurt Cargo Services and Revtrak have already added the AI to their Ops. So, are you going to make the transition? According to Gartner, 40% of large enterprises would have made the transition to AIOps by 2022. If you’re one of them, I recommend you do it for the right reasons, but don’t do it overnight. The transition needs to be gradual and well planned. The first thing you need to do is getting your enterprise data together. If you don’t have sufficient data that’s worthy of analysis, AIOps isn’t going to help you much. Read more: Bridging the gap between data science and DevOps with DataOps.

Machine learning as a service (MLaaS) is an innovation that is growing out of 2 of the most important tech trends - cloud and machine learning. It's significant because it enhances both. It makes cloud an even more compelling proposition for businesses. That's because cloud typically has three major operations: computing, networking and storage. When you bring machine learning into the picture, the data that cloud stores and processes can be used in radically different ways, solving a range of business problems. What is machine learning as a service? Cloud platforms have always competed to be the first or the best to provide new services. This includes platform as a service (PaaS) solutions, infrastructure as a service (IaaS) solutions and software as a service (SaaS) solutions. In essense, cloud providers like AWS and Azure provide sets of software to different things so their customers don't have to. Machine learning as a service is simply another instance of the services offered by cloud providers. It could include a wide range of features, from data visualization to predictive analytics and natural language processing. It makes running machine learning models easy, effectively automating some of the work that might have typically done manually by a data engineering team. Here are the biggest cloud providers who offer machine learning as a service: Google Cloud Platform Amazon Web Services Microsoft Azure IBM Cloud Every platform provides a different suite of services and features. It will ultimately depend on what's most important to you which one you choose. Let's take a look now at the key differences between these cloud providers' machine learning as a service offerings. Comparing the leading MLaaS products Google Cloud AI Google Cloud Platform has always provided their own services to help businesses grow. They provide modern machine learning services with pre-trained models and a service to generate your own tailored models. Majority of Google applications like Photos (image search), the Google app (voice search), and Inbox (Smart Reply) have been built using the same services that they provide to their users. Pros: Cheaper in comparison to other Cloud providers Provides IaaS and PaaS Solutions Cons: Google Prediction API is going to be discontinued (May 1st, 2018) Lacks a visual interface You'll need to know TensorFlow Amazon Machine Learning Amazon Machine Learning provides services for building ML models and generating predictions which help users develop robust, scalable, and cost-effective smart applications. With the help of Amazon Machine Learning you are able to use powerful machine learning technology without having any prior experience in machine learning algorithms and techniques. Pros: Provides versatile automated solutions It's accessible - users don't need to be machine learning experts Cons: The more you use, the more expensive it is Azure Machine Learning Studio Microsoft Azure provides you with Machine Learning Studio - a simple browser-based, drag-and-drop environment which functions without any kind of coding. You are provided with fully-managed cloud services that enable you to easily build, deploy and share predictive analytics solutions. Here you are also provided with a platform (Gallery) to share and contribute to the community. Pros: Consists of most versatile toolset for MLaaS You can contribute to and reuse machine learning solutions from the community Cons: Comparatively expensive A lot of manual work is required Watson Machine Learning Similar to the above platforms, IBM Watson Machine Learning is a service that helps  users to create, train, and deploy self-learning models to integrate predictive capabilities within their applications. This platform provides automated and collaborative workflows to grow intelligent business applications. Pros: Automated workflows Data science skills is not necessary Cons: Comparatively limited APIs and services Lacks streaming analytics Selecting the machine learning as a service solution that's right for you There are so many machine learning as a service solutions out there that it's easy to get confused. The crucial step to take before you make a decision to purchase anything is to plan your business requirements. Think carefully not only about what you want to achieve, but what you already do too. You want your MLaaS solution to easily integrate into the way you currently work. You also don't want it to replicate any work you're currently doing that you're pretty happy with. It gets repeated so much but it remains as true as it has ever been - make sure your software decisions are fully aligned with your business needs. It's easy to get seduced by the promise of innovative new tools, but without the right alignment they're not going to help you at all.

This year, IBM Think 2018 was hosted in Las Vegas from March 20 to 22. It was one of the most anticipated IBM events in 2018, with over 40,000 developers as well as technology and business leaders in attendance. Considered IBM’s flagship conference, Think 2018 combined previous conferences such as IBM InterConnect and World of Watson. IBM Think 2018: Key Takeaways IBM Watson Studio announced - A platform where data professionals in different roles can come together and build end-to-end Artificial Intelligence workflows Integration of IBM Watson with Apple's Core ML, for incorporating custom machine learning models into iOS apps IBM Blockchain platform announced, for Blockchain developers to build enterprise-grade decentralized applications Deep Learning as a Service announced as a part of the Watson Studio, allowing you to train deep learning models more efficiently Fabric for Deep Learning open-sourced, so that you can use the open source deep learning framework to train your models and then integrate them with the Watson Studio Neural Network Modeler announced for Watson Studio, a GUI tool to design neural networks efficiently, without a lot of manual coding IBM Watson Assistant announced, an AI-powered digital assistant, for automotive vehicles and hospitality Here are some of the announcements and key takeaways which have excited us, as well as the developers all around the world! IBM Watson Studio announced One of the biggest announcements of the event was the IBM Watson Studio - a premier tool that brings together data scientists, developers and data engineers to collaborate, build and deploy end-to-end data workflows. Right from accessing your data source to deploying accurate and high performance models, this platform does it all. It is just what enterprises need today to leverage Artificial Intelligence in order to accelerate research, and get intuitive insights from their data. IBM Watson Studio's Lead Product Manager, Armand Ruiz, gives a sneak-peek into what we can expect from Watson Studio. Collaboration with Apple Core ML IBM took their relationship with Apple to another level by announcing their collaboration to develop smarter iOS applications. IBM Watson’s Visual Recognition Service can be used to train custom Core ML machine learning models, which can be directly used by iOS apps. The latest announcement at IBM Think 2018 comes as no surprise to us, considering IBM had released new developer tools for enterprise development using the Swift language. IBM Watson Assistant announced IBM Think 2018 also announced the evolution of Watson Conversation to Watson Assistant, introducing new features and capabilities to deliver a more engaging and personalized customer experience. With this, IBM plans to take the concept of AI assistants for businesses on to a new level. Currently in the beta program, there are 2 domain-specific solutions available for use on top of Watson Assistant - namely Watson Assistant for Automotive and Watson Assistant for Hospitality. IBM Blockchain Platform Per Juniper Research, more than half of the world’s big corporations are considering adoption of or are already in the process of adopting Blockchain technology. This presents a serious opportunity for a developer centric platform that can be used to build custom decentralized networks. IBM, unsurprisingly, has identified this opportunity and come up with a Blockchain development platform of their own - the IBM Blockchain Platform. Recently launched as a beta, this platform offers a pay-as-you-use option for Blockchain developers to develop their own enterprise-grade Blockchain solutions without any hassle. Deep Learning as a Service Training a deep learning model is quite tricky, as it requires you to design the right kind of neural networks along with having the right hyperparameters. This is a significant pain point for the data scientists and machine learning engineers. To tackle this problem,  IBM announced the release of Deep Learning as a Service as part of the Watson Studio. It includes the Neural Network Modeler (explained in detail below) to simplify the process of designing and training neural networks. Alternatively, using this service, you can leverage popular deep learning libraries and frameworks such as PyTorch, Tensorflow, Caffe, Keras to train your neural networks manually. In the process, IBM also open sourced the core functionalities of Deep Learning as a Service as a separate project - namely Fabric for Deep Learning. This allows models to be trained using different open source frameworks on Kubernetes containers, and also make use of the GPUs’ processing power. These models can then eventually be integrated to the Watson Studio. Accelerating deep learning with the Neural Network Modeler In a bid to reduce the complexities and the manual work that go into designing and training neural networks, IBM introduced a beta release of the Neural Network Modeler within the Watson Studio. This new feature allows you to design and model standardized neural network models without going into a lot of technical details, thanks to its intuitive GUI. With this announcement, IBM aims to accelerate the overall process of deep learning, so that the data scientists and machine learning developers can focus on the thinking more than operational side of things. At Think 2018, we also saw the IBM Research team present their annual ‘5 in 5’ predictions. This session highlighted the 5 key innovations that are currently in research, and are expected to change our lives in the near future. With these announcements, it’s quite clear that IBM are well in sync with the two hottest trends in the tech space today - namely Artificial Intelligence and Blockchain. They seem to be taking every possible step to ensure they’re right up there as the preferred choice of tool for data scientists and machine learning developers. We only expect the aforementioned services to get better and have more mainstream adoption with time, as most of these services are currently in the beta stage. Not just that, there’s scope for more improvements and addition of newer functionalities as they develop these platforms. What did you think of these announcements by IBM? Do let us know!

IBM’s mission has always been to innovate and in the process, change the way the world operates. With this objective in mind, IBM Research started a conversation termed as ‘5 in 5’ way back in 2012, giving their top 5 predictions every year at IBM Think 2018 on how technology would change the world. These predictions are usually the drivers for their research and innovation - and eventually solving the problems by coming up with efficient solutions to them. Here are the 5 predictions made by IBM Research for 2018: More secure Blockchain products: In order to avoid counterfeit Blockchain products, the technology will be coupled with cryptographic solutions to develop decentralized solutions. Digital transactions are often subject to frauds, and securing them with crypto-anchors is seen as the way to go forward. Want to know how this can be achieved? You might want to check out IBM’s blog on crypto-anchors and their real world applications. If you are like me, you’d rather watch IBM researcher Andres Kind explain what crypto-anchors are in a fast paced science slam session. Sophisticated cyber attacks will continue to happen: Cyber attacks resulting in the data leaks or stealing of confidential data is not news to us. The bigger worry, though, is that the current methodologies to prevent these attacks are not proving to be good enough. IBM predicts this is only going to get worse, with more advanced and sophisticated cyber attacks breaking into the current secure systems with ease. IBM Research also predicted the rise of ‘lattice cryptography’, a new security mechanism offering a more sophisticated layer of protection for the systems. You can read more about lattice cryptography technology on IBM’s official blog. Or, you can watch IBM researcher Cecilia Boschini explain what is lattice cryptography in 5 minutes on one of IBM’s famous science slam sessions. Artificial Intelligence-powered bots will help clean the oceans: Our marine ecosystem seems to be going from bad to worse. This is mainly due the pollution and toxic wastes being dumped into it. IBM predicts that AI-powered autonomous bots, deployed and controlled on the cloud, can help relieve this situation by monitoring the water bodies for water quality and pollution levels. You can learn more about how these autonomous bots will help save the seas in this interesting talk by Tom Zimmerman. An unbiased AI system: Artificially designed systems  are only as good as the data being used to build them. This data may be impure, or may contain flaws or bias pertaining of color, race, gender and so on. Going forward, new models which mitigate these biases and ensure more standard, bias-free predictions will be designed. With these models, certain human values and principles will be considered for effective decision-making. IBM researcher Francesca Rossi talks about bias in AI and the importance of building fair systems that help us make better decisions. Quantum Computing will go mainstream: IBM predicts that quantum computing will get out of research labs and gain mainstream adoption in the next 5 years. Problems considered to be difficult or unsolvable today due to their sheer scale or complexity can be tackled with the help of quantum computing. To know more, let IBM researcher Talia Gershon take you through the different aspects of quantum computing and why it is expected to be a massive hit. Amazingly, most of the predictions from the past have turned out to be true. For instance, IBM predicted the rise of Computer Vision technology in 2012, where computers would be able to not only process images, but also understand their ‘features’. It remains to be seen how true this year’s predictions will turn out to be. However, considering the rate at which the research on AI and other tech domains is progressing and being put to practical use, we won’t be surprised if they all become a reality soon. What do you think?

Amazon Sagemaker was launched by Amazon back in November 2017. It was built with the promise of simplifying machine learning on the cloud. The software was a response not only to the increasing importance of machine learning, but also the fact that there is a demand to perform machine learning in the cloud. Amazon Sagemaker is clearly a smart move by Amazon that will consolidate the dominance of AWS in the cloud market. What is Amazon Sagemaker? Amazon Sagemaker is Amazon’s premium cloud-based service which serves as a platform for machine learning developers and data scientists to build, train and deploy machine learning models on the cloud. One of the features that makes Sagemaker stand out from the rest is that it is business-ready. This means machine learning models can be optimized for high performance and deployed at scale to work on data with varying sizes and complexity. The basic intention of Sagemaker, as Vogels mentioned in his keynote, is to remove any barriers that slow down the machine learning process for developers. In a standard machine learning process, a developer spends most of the time doing the following standard tasks: Collecting, cleaning and preparing the training data set. Selecting the most appropriate algorithm for the machine learning problem Training the model for accurate prediction Optimizing the model’s performance Integrating the model with the application Deploying the application to production Most of these tasks require a lot of expertise, and more importantly, time and efforts. Not to mention the computational resources such as storage space and processing memory. The larger the dataset, the bigger this problem becomes. Amazon Sagemaker removes these complexities by providing a solid platform with built-in modules that can be used together or individually to complete each of the above tasks with relative ease. How Amazon Sagemaker Works Amazon Sagemaker offers a lot of options for machine learning developers to train and optimize their machine learning models to work at scale. For starters, Sagemaker comes integrated with hosted Jupyter notebooks to allow developers to visually explore and analyze their dataset. You can also move your data directly from popular Amazon databases such as RDS, DynamoDB and Redshift into S3 and conduct your analysis there. The simple block diagram below demonstrates the core working of Amazon Sagemaker: Amazon Sagemaker includes 12 high performance, production-ready algorithms which can be used to build and deploy models at scale. Some of the popular ones include k-means clustering, Principal Component Analysis (PCA), neural topic modeling, and more. It comes pre-configured with popular machine learning and deep learning frameworks such as Tensorflow, PyTorch, Apache MXNet and more, but you can also use your own framework without any hassle. Once your model is trained, Sagemaker makes use of the AWS’ auto-scaled clusters to deploy the model, making sure the model doesn’t lack in performance and is highly available at all times. Not just that, Sagemaker also includes built-in testing capabilities for you to test and check your model for any issues, before it can be deployed for production. Benefits of using Amazon Sagemaker Business are likely to adopt Amazon Sagemaker, mainly because of the fact that it makes the whole machine learning process so effortless. With Sagemaker, it becomes very easy to build and deploy smarter applications that give accurate predictions, and thereby help increase the business profitability. Significantly reduces time: With built-in modules, Sagemaker significantly reduces the time required to do a variety of machine learning tasks, and the models can be deployed to production in very little time. This is important for businesses, as near-real time insights obtained from smart applications help them optimize their processes quickly, and effectively get an edge over their competition. Effortless and more productive machine learning: By virtue of the one-click training and deployment feature offered by Sagemaker, machine learning engineers and developers can now focus on asking the right questions of the data, and focus on the results rather than the process. They can also devote more time to optimizing the model rather than focusing on collecting and cleaning the data, which takes up most of their time. Flexibility in using the algorithms and frameworks: With Sagemaker, developers have the freedom to choose the best-possible algorithm and tool for performing machine learning effectively. Easy integration, access and optimization: The models trained using Sagemaker can be integrated into an existing business application seamlessly, and are optimized for speed and high performance. Backed by the computational power of AWS, business can rest assured their applications will continue to perform optimally without any risk of failure. Sagemaker - Amazon’s answer to Cloud Auto ML In a 3-way cloud war between Google, Microsoft and Amazon, it is clear Google and Amazon are trying to go head to head in order to establish their supremacy in the market, especially in the AI space. Sagemaker is Amazon’s answer to Google’s Cloud Auto ML, which was made publicly available in January, and delivers a similar promise - making machine learning easier than ever for developers. With Amazon serving a large customer-base, a platform like Sagemaker helps them to create a system that runs at scale and handles vast amounts of data quite effortlessly.  Amazon is yet to release any technical paper on how Sagemaker’s streaming algorithms work, but that will certainly be something to look out for in the near future. Considering Amazon identifies AI as key to their future product development, to think of Sagemaker as a better, more complete cloud service which also has deep learning capabilities is definitely not far-fetched.

The adoption of Computer Vision has been steadily picking up pace over the past decade, but there’s been a spike in adoption of various computer vision tools in recent times, thanks to its implementation in fields like IoT, manufacturing, healthcare, security, etc. Computer vision tools have evolved over the years, so much so that computer vision is now also being offered as a service. Moreover, the advancements in hardware like GPUs, as well as machine learning tools and frameworks make computer vision much more powerful in the present day. Major cloud service providers like Google, Microsoft and AWS have all joined the race towards being the developers’ choice. But which tool should you choose? Today I’ll take you through a list of the top tools and will help you understand which one to pick up, based on your need. Computer Vision Tools/Libraries OpenCV: Any post on computer vision is incomplete without the mention of OpenCV. OpenCV is a great performing computer vision tool and it works well with C++ as well as Python. OpenCV is prebuilt with all the necessary techniques and algorithms to perform several image and video processing tasks. It’s quite easy to use and this makes it clearly the most popular computer vision library on the planet! It is multi-platform, allowing you to build applications for Linux, Windows and Android. At the same time, it does have some drawbacks. It gets a bit slow when working through massive data sets or very large images. Moreover, on its own, it doesn’t have GPU support and relies on CUDA for GPU processing. Matlab: Matlab is a great tool for creating image processing applications and is widely used in research. The reason being that Matlab allows quick prototyping. Another interesting aspect is that Matlab code is quite concise, as compared to C++, making it easier to read and debug. It tackles errors before execution by proposing some ways to make the code faster. On the downside, Matlab is a paid tool. Also, it can get quite slow during execution time, if that’s something that concerns you much. Matlab is not your go to tool in an actual production environment, as it was basically built for prototyping and research. AForge.NET/Accord.NET: You’ll be excited to know that image processing is possible even if you’re a C# and .NET developer, thanks to AForge/Accord. It’s a great tool that has a lot of filters and is great for image manipulation and different transforms. The Image Processing Lab allows for filtering capabilities like edge detection and more. AForge is extremely simple to use as all you need to do is adjust parameters from a user interface. Moreover, its processing speeds are quite good. However, AForge doesn’t possess the power and capabilities of other tools like OpenCV, like advanced motion picture analysis or even advanced processing on images. TensorFlow: TensorFlow has been gaining popularity over the past couple of years, owing to its power and ease of use. It lets you bring the power of Deep Learning to computer vision and has some great tools to perform image processing/classification - it’s API-like graph tensor. Moreover, you can make use of the Python API to perform face and expression detection. You can also perform classification using techniques like regression. Tensorflow also allows you to perform computer vision of tremendous magnitudes. One of the main drawbacks of Tensorflow is that it’s extremely resource hungry and can devour a GPU’s capabilities in no time, quite uncalled for. Moreover, if you wanted to learn how to perform image processing with TensorFlow, you’d have to understand what Machine and Deep Learning is, write your own algorithms and then go forward from there. CUDA: CUDA is a platform for parallel computing, invented by NVIDIA. It enables great boosts in computing performance by leveraging the power of GPUs. The CUDA Toolkit includes the NVIDIA Performance Primitives library which is a collection of signal, image, and video processing functions. If you have large images to process, that are GPU intensive, you can choose to use CUDA. CUDA is easy to program and is quite efficient and fast. On the downside, it is extremely high on power consumption and you will find yourself reformulating for memory distribution in parallel tasks. SimpleCV: SimpleCV is a framework for building computer vision applications. It gives you access to a multitude of computer vision tools on the likes of OpenCV, pygame, etc. If you don’t want to get into the depths of image processing and just want to get your work done, this is the tool to get your hands on. If you want to do some quick prototyping, SimpleCV will serve you best. Although, if your intention is to use it in heavy production environments, you cannot expect it to perform on the level of OpenCV. Moreover, the community forum is not very active and you might find yourself running into walls, especially with the installation. GPUImage: GPUImage is a framework or rather, an iOS library that allows you to apply GPU-accelerated effects and filters to images, live motion video, and movies. It is built on OpenGL ES 2.0. Running custom filters on a GPU calls for a lot of code to set up and maintain. GPUImage cuts down on all of that boilerplate and gets the job done for you. Computer Vision as a Service: Google Cloud and Mobile Vision APIs: Google Cloud Vision API enables developers to perform image processing by encapsulating powerful machine learning models in a simple REST API that can be called in an application. Also, its Optical Character Recognition (OCR) functionality enables you to detect text in your images. The Mobile Vision API lets you detect objects in photos and video, using real-time on-device vision technology. It also lets you scan and recognise barcodes and text. Amazon Rekognition: Amazon Rekognition is a deep learning-based image and video analysis service that makes adding image and video analysis to your applications, a piece of cake. The service can identify objects, text, people, scenes and activities, and it can also detect inappropriate content, apart from providing highly accurate facial analysis and facial recognition for sentiment analysis. Microsoft Azure Computer Vision API: Microsoft’s API is quite similar to its peers and allows you to analyse images, read text in them, and analyse video in near-real time. You can also flag adult content, generate thumbnails of images and recognise handwriting. Bonus: SciPy and NumPy: I thought I’d add these in as well, since I’ve seen quite a few developers use Python to build computer vision applications (without OpenCV, that is). SciPy and NumPy are quite powerful enough to perform image processing. scikit-image is a Python package that is dedicated towards image processing, which uses native NumPy and SciPy arrays as image objects. Moreover, you get to use the cool IPython interactive computing environment and you can also choose to include OpenCV if you want to do some more hardcore image processing. Well there you have it, these were the top tools for computer vision and image processing. Head on over and check out these resources, to get working with some of the top tools used in the industry.

DeepMind has always open sourced their projects with a bang. Last year, it announced that it is going to open source Sonnet, a library for quickly building neural network modules with Tensorflow. Deepmind shifted from Torch to Tensorflow as their choice of framework since early 2016, after it was acquired by Google in 2014. Why Sonnet if you have TensorFlow? Since adopting TensorFlow as the choice of their framework, DeepMind has enjoyed the flexibility and adaptiveness of TF for building higher-level frameworks. In order to build neural network modules with Tensorflow, they created a framework called Sonnet. Sonnet doesn’t typically replace TensorFlow; it just eases the process of constructing neural networks. Prior to Sonnet, DeepMind developers were forced to become intimately familiar with the underlying TensorFlow graphs in order to correctly architect its applications. With Sonnet, the creation of neural network components is quite easy as it first constructs Python objects which represent some part of a neural network, and then separately connect these objects into the TensorFlow computation graph. What makes Sonnet special? Sonnet uses Modules. Modules encapsulate elements of a neural network which in turn abstracts low-level aspects of TensorFlow applications. Sonnet enables developers to build their own Modules using a simple programming model. These Modules simplify the neural network training and can help to implement individual neural networks that can be combined to implement higher-level networks. Developers can also easily extend Sonnet by implementing their own modules. Using Sonnet, it becomes easier to switch between different models, allowing engineers to freely conduct experiments without worrying about hampering their entire projects. Why open source Sonnet? The announcement of Sonnet open sourcing came on April 7, 2017. Most people appreciated it as a move in the right direction. One of the focal purpose of DeepMind to open source Sonnet was to make the developer community to use Sonnet to take their own research forwards.  According to FossBytes, "DeepMind foresees Sonnet to be used by the community as a research propellant." With this open sourcing, the machine learning community can then more actively contribute back by utilizing Sonnet in their own projects. Moreover, if the community becomes accustomed and acquainted with DeepMind’s internal libraries, it will become easier for the DeepMind group to release other Machine learning models alongside research papers. Certain experienced developers also point out that using TensorFlow and Sonnet together is similar to using TensorFlow and Torch together, with a Reddit comment stating “DeepMind's trying to turn TensorFlow into Torch”. Nevertheless, open sourcing of Sonnet is seen as DeepMind’s part of their broader commitment to open source AI research. Also, as Sonnet is adopted by the community more similar frameworks are also likely to develop that make neural network construction easier using TensorFlow as the underlying runtime. Taking a further step towards democratization of machine learning and its subsidies. Sonnet is already available on GitHub and will be regularly updated by the DeepMind team to match the in-house version.

What is Emoji Scavenger Hunt? Emoji Scavenger Hunt is a game built using neural networks. Developed by Google using TensorFlow.js, a version of the machine learning library designed to run on browsers, the game showcases how machine learning can be brought to web applications. But more importantly, TensorFlow.js, which was announced at the end of March at the TensorFlow Developer Summit looks like it could be a tool to define the next few years of web development, making machine learning more accessible to JavaScript developers than ever before. Start playing now. At the moment Emoji Scavenger Hunt is pretty basic, but the central idea is pretty cool. When you open up the web page in your browser and click 'Let's Play', the app asks for access to your camera. The game then starts: you'll see a countdown, before your camera opens and the web application asks you to find an example of an emoji in the real world. If you find yourself easily irritated you're probably not going to get addicted, as Google seem to have done their best to cultivate an emoji-esque mise en scene. But the game nevertheless highlights not only how neural networks work, but also, in the context of TensorFlow.js, how they might operate in a browser. Of course, one of the reasons Emoji Scavenger Hunt is so basic is because a core part of the game is training the neural network. Presumably, as more people play it, the neural network will improve at 'guessing' what objects in the real world relate to which emoji on your keyboard. TensorFlow.js will bring machine learning to the browser What's exciting is how TensorFlow.js might help shape the future of web development. It's going to make it much easier for JavaScript developers to get started with machine learning - on Reddit a number of users were thankful that they could now use TensorFlow without touching a line of Python code. On the other hand - perhaps a little less likely - TensorFlow.js might lead to more machine learning developers using JavaScript. If games like Emoji Scavenger Hunt become the norm, engineers and data scientists will have a new way to train algorithms - getting users to do it for them. TensorFlow.js and deeplearn.js Eagle-eyed readers who have been watching TensorFlow closely might be thinking here - what about deeplearn.js? Fortunately, the TensorFlow team have an answer: TensorFlow.js... is the successor to deeplearn.js, which is now called TensorFlow.js Core. TensorFlow.js and the future of machine learning The announcement of TensorFlow.js highlights that Google and the core development team behind TensorFlow have a clear focus on the future. They're already the definitive library for machine learning and deep learning. What this will do is spread its dominance into new domains. Emoji Scavenger Hunt is pointing the way - we're sure to see plenty of machine learning imitators and innovators over the next few years.

The ICLR 2018 accepted paper, i-RevNet: Deep Invertible Networks, introduces i-RevNet, an invertible convolutional network, that does not discard any information about the input while classifying images. This paper is authored by Jörn-Henrik Jacobsen, Arnold W.M. Smeulders, and Edouard Oyallon. The 6th annual ICLR conference is scheduled to happen between April 30 - May 03, 2018. i-RevNet, a deep invertible convolutional network What problem is the paper attempting to solve? A CNN is generally composed of a cascade of linear and nonlinear operators. These operators are very effective in classifying images of all sorts but reveal little information about the contribution of the internal representation to the classification. The learning process of a CNN works by a regular reduction of large amounts of uninformative variability in the images to reveal the essence of the visual class. However, the extent to which information is discarded is lost somewhere in the intermediate nonlinear processing steps. Also, there is a wide belief, that discarding information is essential for learning representations that generalize well to unseen data. The authors of this paper show that discarding information is not necessary and propose to explain this theory with empirical evidence. This paper also provides an understanding of the variability reduction process by proposing an invertible convolutional network. The i-RevNet does not discard any information about the input while classifying images. It has a built-in pseudo-inverse, allowing for easy inversion.  It basically uses linear and invertible operators for performing downsampling, instead of non-invertible variants like spatial pooling. Paper summary i-RevNet is an invertible deep network, which builds upon the recently introduced RevNet, where the non-invertible components of the original RevNets are replaced by invertible ones. i-RevNets retain all information about the input signal in any of their intermediate representations up until the last layer. They achieve the same performance on Imagenet compared to similar non-invertible RevNet and ResNet architectures. The above image describes the blocks of an i-RevNet. The strategy implemented by an i-RevNet consists in an alternation between additions, and nonlinear operators, while progressively down-sampling the signal operators. The pair of the final layer is concatenated through a merging operator. Using this architecture, the authors avoid the non-invertible modules of a RevNet (e.g. max-pooling or strides) which are necessary to train them in a reasonable time and are designed to build invariance w.r.t. Translation variability. Their method replaces the non-invertible modules by linear and invertible modules Sj, that can reduce the spatial resolution while maintaining the layer’s size by increasing the number of channels. Key Takeaways This work provides a solid empirical evidence that learning invertible representations does not discard any information about their input on large-scale supervised problems. i-RevNet, the invertible network proposed, is a class of CNN which is fully invertible and permits to exactly recover the input from its last convolutional layer. i-RevNets achieve the same classification accuracy in the classification of complex datasets as illustrated on ILSVRC-2012 when compared to the RevNet and ResNet architectures with a similar number of layers. The inverse network is obtained for free when training an i-RevNet, requiring only minimal adaption to recover inputs from the hidden representations. Reviewer feedback summary Overall Score: 25/30 Average Score: 8.3 Reviewers agreed the paper is a strong contribution, despite some comments about the significance of the result; i.e., why is invertibility a "surprising" property for learnability, in the sense that F(x) = {x,  phi(x)}, where phi is a standard CNN satisfies both properties: invertible and linear measurements of F producing good classification. Having said that, the reviews agreed that the paper is well written and easy to follow and considered it to be a great contribution to the ICLR conference.