Summary of Fine-grained Background Representation For Weakly Supervised Semantic Segmentation, by Xu Yin et al.
Fine-grained Background Representation for Weakly Supervised Semantic Segmentation
by Xu Yin, Woobin Im, Dongbo Min, Yuchi Huo, Fei Pan, Sung-Eui Yoon
First submitted to arxiv on: 22 Jun 2024
Categories
- Main: Computer Vision and Pattern Recognition (cs.CV)
- Secondary: Artificial Intelligence (cs.AI)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary The proposed method for generating reliable pseudo masks from image-level labels in the weakly supervised semantic segmentation task addresses the challenge of discriminating foreground objects from suspicious background pixels. The fine-grained background representation (FBR) method abandons traditional class prototypes and pixel-level features, instead using a novel negative region of interest (NROI) to capture fine-grained background semantics. This is combined with an active sampling strategy for mining FG negatives and intra-foreground contrastive learning to activate object regions. The proposed method achieves state-of-the-art results under various WSSS settings across benchmarks, leveraging solely image-level labels as supervision. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper proposes a new way to generate masks from images without using much information about what’s in the image. This is important because it can help machines learn to do tasks like segmentation and object detection better. The method uses something called fine-grained background representation (FBR) that helps identify what’s in the background of an image. It also includes a way to find negative examples, which are important for training machine learning models. The results show that this method works well and can be used with different models. |
Keywords
» Artificial intelligence » Machine learning » Object detection » Semantic segmentation » Semantics » Supervised
