Segment objects by learning a stable hardness value for pixels with Hardness-Level-Learning
Segment objects by learning a stable hardness value for pixels with Hardness-Level-Learning
Not All Pixels Are Equal: Learning Pixel Hardness for Semantic Segmentation
arXiv paper abstract https://arxiv.org/abs/2305.08462
arXiv PDF paper https://arxiv.org/pdf/2305.08462.pdf
GitHub https://github.com/Menoly-xin/Hardness-Level-Learning
Semantic segmentation … in some hard areas (e.g., small objects or thin parts) is still not promising.
… existing hard pixel mining … rely on … loss value, which … decrease during training … Intuitively … hardness … depends on image structure and is expected to be stable.
… propose to learn pixel hardness for semantic segmentation, leveraging hardness information contained in global and historical loss values.
… add a gradient-independent branch for learning a hardness level (HL) map … encourages large hardness values in difficult areas, leading to appropriate and stable HL map.
… proposed method can be applied to most segmentation methods with no and marginal extra cost during inference and training, respectively.
… method achieves consistent/significant improvement (1.37% mIoU on average) over most popular semantic segmentation methods on Cityscapes dataset, and demonstrates good generalization ability across domains …
Stay up to date. Subscribe to my posts https://morrislee1234.wixsite.com/website/contact
Web site with my other posts by category https://morrislee1234.wixsite.com/website
LinkedIn https://www.linkedin.com/in/morris-lee-47877b7b
