Summary of Dual Mean-teacher: An Unbiased Semi-supervised Framework For Audio-visual Source Localization, by Yuxin Guo et al.
Dual Mean-Teacher: An Unbiased Semi-Supervised Framework for Audio-Visual Source Localization
by Yuxin Guo, Shijie Ma, Hu Su, Zhiqing Wang, Yuhao Zhao, Wei Zou, Siyang Sun, Yun Zheng
First submitted to arxiv on: 5 Mar 2024
Categories
- Main: Computer Vision and Pattern Recognition (cs.CV)
- Secondary: Machine Learning (cs.LG); Multimedia (cs.MM); Sound (cs.SD); Audio and Speech Processing (eess.AS)
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 Dual Mean-Teacher (DMT) framework is a novel semi-supervised learning method for Audio-Visual Source Localization (AVSL). Unlike existing self-supervised contrastive learning approaches, DMT uses two teachers pre-trained on limited labeled data to filter out noisy samples and generate high-quality pseudo-labels. By leveraging both labeled and unlabeled data, DMT outperforms state-of-the-art methods by a large margin, with improved precision and recall. The framework is demonstrated on several AVSL datasets, achieving significant improvements over existing methods. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary AVSL aims to find sounding objects in videos given paired audio clips. Existing methods struggle to achieve precise localization, especially for small objects, due to the lack of bounding-box annotations. This paper proposes a new semi-supervised learning framework called Dual Mean-Teacher (DMT). DMT uses two teachers pre-trained on limited labeled data to filter out noisy samples and generate high-quality pseudo-labels. This helps to utilize both labeled and unlabeled data more effectively, leading to improved results. |
Keywords
* Artificial intelligence * Bounding box * Precision * Recall * Self supervised * Semi supervised
