Summary of Ss-ada: a Semi-supervised Active Domain Adaptation Framework For Semantic Segmentation, by Weihao Yan et al.
SS-ADA: A Semi-Supervised Active Domain Adaptation Framework for Semantic Segmentation
by Weihao Yan, Yeqiang Qian, Yueyuan Li, Tao Li, Chunxiang Wang, Ming Yang
First submitted to arxiv on: 17 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 semi-supervised active domain adaptation (SS-ADA) framework for semantic segmentation integrates active learning into semi-supervised semantic segmentation, allowing it to achieve the accuracy of supervised learning with a limited amount of labeled data from the target domain. The framework employs an image-level acquisition strategy and an IoU-based class weighting strategy to alleviate the class imbalance problem using annotations from active learning. Experiments on synthetic-to-real and real-to-real domain adaptation settings demonstrate the effectiveness of SS-ADA, which can achieve or even surpass the accuracy of its supervised learning counterpart with only 25% of the target labeled data when using a real-time segmentation model. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary Semantic segmentation is crucial for intelligent vehicles, but labeling data for new driving scenarios is time-consuming and expensive. Semi-supervised methods have been proposed to use unlabeled images, but they still fall short of accuracy required for practical applications. A novel approach called SS-ADA combines active learning with semi-supervised semantic segmentation to achieve high accuracy with limited labeled data. The framework uses an image-level strategy to select which images to label and adjusts the class weights based on how well each class is performing. |
Keywords
» Artificial intelligence » Active learning » Domain adaptation » Semantic segmentation » Semi supervised » Supervised
