Semantic Segmentation and Matting

Introduction

1. What is Semantic Segmentation and its variations?
   • Semantic Segmentation, instance segmentation, pan-optic segmentation
   • Open Vocabulary settings
2. What is matting? How is it distinct from Semantic Segmentation?
3. What are the popular loss functions for precise semeantic segmentation?
4. What are the popular loss functions for matting?
5. What are the popular benchmarks?
6. How do people collect ground truth for segmentation and matting?
   • How do we convert subjective ratings into objective metrics?
   • What are the computational challenges involved?
7. What are the important applications of matting and sementic segmentation?

References

1. https://luozm.github.io/cv-tasks
2. https://www.v7labs.com/blog/panoptic-segmentation-guide
3. https://github.com/ngthanhtin/owlvit_segment_anything
4. https://github.com/JunMa11/SegLoss
5. https://lars76.github.io/2018/09/27/loss-functions-for-segmentation.html
6. https://github.com/JunMa11/SegWithDistMap
7. Deep image matting
8. https://github.com/JestelleWang/MVIG-training-PS/blob/master/Papers/What%27s%20the%20Role%20of%20Image%20Matting%20in%20Image%20Segmentation%3F.md