Summary of Resilient Practical Test-time Adaptation: Soft Batch Normalization Alignment and Entropy-driven Memory Bank, by Xingzhi Zhou et al.
Resilient Practical Test-Time Adaptation: Soft Batch Normalization Alignment and Entropy-driven Memory Bank
by Xingzhi Zhou, Zhiliang Tian, Ka Chun Cheung, Simon See, Nevin L. Zhang
First submitted to arxiv on: 26 Jan 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 proposes a new test-time domain adaptation method called ResiTTA that aims to improve model performance in real-world scenarios where the target domain changes continuously and test samples are non-i.i.d. The method focuses on parameter resilience and data quality by developing a resilient batch normalization technique and an entropy-driven memory bank. This framework periodically adapts the source domain model using a teacher-student model, incorporating soft alignment losses on batch normalization. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary The ResiTTA method is designed to help models adapt better to changing target domains and non-i.i.d. test samples, which can be a problem in real-world scenarios. The approach uses a combination of techniques, including resilient batch normalization and an entropy-driven memory bank, to improve model performance. The framework also incorporates soft alignment losses on batch normalization to further adapt the source domain model. |
Keywords
* Artificial intelligence * Alignment * Batch normalization * Domain adaptation * Student model
