Part 5:Current Challenges and Future Perspectives

In this part, you learn about a well-known issue that usually limits the usability of synthetic data. You will learn about the domain gap problem and why diversity and photorealism are some of the main challenges toward generating useful and large-scale synthetic data in practice. You will learn about various approaches to bridge the domain gap and improve the diversity and photorealism of your synthetic data. Then, we will recap the benefits of synthetic data-based solutions, challenges, and limitations. Finally, we will highlight some interesting future perspectives.

This part has the following chapters:

• Chapter 14, Synthetic-to-Real Domain Adaptation
• Chapter 15, Diversity Issues in Synthetic Data
• Chapter 16, Photorealism in Computer Vision
• Chapter 17, Conclusion