What's Next - Wrapping Up and Future Directions

The use of IoT will be come omnipresent in our life through various applications, including connected healthcare, smart home, and smart city. Although IoT applications are exciting, there are significant scientific and technological challenges to be overcome before they can be fully realized. One of the key technical challenges for IoT is the design and development of Deep Learning (DL) models that work well in resource-constrained IoT devices or edge/fog computing devices, and those that meet the real-time response requirements. In this final chapter, we shall first present a summary of the earlier chapters, and then use examples to discuss the main challenges faced by the existing DL techniques in their development and implementation of resource-constrained and embedded IoT environments. Finally, we will summarize a few existing...