Conclusion and Future Directions

In the final chapter, we will present a concise review of the content covered in this book, encapsulating different aspects explored in the previous chapters. Furthermore, this chapter will also look into the future of deep-learning-based video understanding by briefly discussing several promising directions, e.g., the construction of large-scale video foundation models, the application of large language models (LLMs) in video understanding, etc. By depicting these exciting prospects, we encourage the readers to embark on new endeavors to contribute to the advancement of this field.