Summary of One-step Noisy Label Mitigation, by Hao Li et al.
One-step Noisy Label Mitigation
by Hao Li, Jiayang Gu, Jingkuan Song, An Zhang, Lianli Gao
First submitted to arxiv on: 2 Oct 2024
Categories
- Main: Computer Vision and Pattern Recognition (cs.CV)
- Secondary: Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
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 A new noise mitigation method for large-scale pre-training tasks is proposed, which addresses the limitations of existing methods by exploiting high-dimensional orthogonality to identify a robust boundary in cone space. The One-step Anti-Noise (OSA) paradigm employs an estimator model and scoring function to assess the noise level of input pairs through one-step inference, making it cost-efficient and easy to deploy. OSA is shown to be effective across various benchmarks, models, and tasks, improving training robustness, task transferability, and reducing computational costs. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper introduces a new way to reduce mistakes in machine learning models caused by noisy labels. The approach uses special properties of high-dimensional spaces to identify clean and noisy data points. It then proposes an algorithm called One-step Anti-Noise (OSA) that can be used with any machine learning model to improve the quality of training and make it more robust. This method is faster, easier to use, and performs better than existing methods. |
Keywords
» Artificial intelligence » Inference » Machine learning » Transferability
