Summary of Weakly-supervised Residual Evidential Learning For Multi-instance Uncertainty Estimation, by Pei Liu and Luping Ji
Weakly-Supervised Residual Evidential Learning for Multi-Instance Uncertainty Estimation
by Pei Liu, Luping Ji
First submitted to arxiv on: 7 May 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: None
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 paper addresses the challenge of uncertainty estimation (UE) in scenarios where there is no sufficiently-labeled data to support fully-supervised learning. Specifically, it focuses on Multi-Instance UE (MIUE), which often has only weak instance annotations. The proposed baseline scheme, Multi-Instance Residual Evidential Learning (MIREL), jointly models the high-order predictive distribution at bag and instance levels for MIUE. MIREL outperforms representative UE methods by large margins, especially in instance-level UE tasks. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary Uncertainty estimation is important for making safe decisions, but it’s hard when we don’t have enough labeled data to learn from. In this case, the paper shows how to use weak labels to improve uncertainty estimation. It introduces a new method called MIREL that works well even with just weak instance annotations. The results show that MIREL can do better than other methods in some cases. |
Keywords
» Artificial intelligence » Supervised
