Summary of A Systematic Review on Long-tailed Learning, by Chongsheng Zhang et al.
A Systematic Review on Long-Tailed Learning
by Chongsheng Zhang, George Almpanidis, Gaojuan Fan, Binquan Deng, Yanbo Zhang, Ji Liu, Aouaidjia Kamel, Paolo Soda, João Gama
First submitted to arxiv on: 1 Aug 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Artificial Intelligence (cs.AI); Computer Vision and Pattern Recognition (cs.CV); Multimedia (cs.MM)
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 This paper presents a comprehensive survey on recent advances in long-tailed visual learning, which aims to develop high-performance models for datasets with imbalanced distributions. The authors propose a new taxonomy of eight dimensions for long-tailed learning, including data balancing, neural architecture, and loss function adjustments. They then review existing methods and analyze their differences from imbalance learning approaches. This work provides an overview of the latest advances in this field, shedding light on the commonalities and alignable differences between various methods. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary Long-tailed data is a special type of imbalanced data with many minority classes that have a big impact. Long-tailed learning tries to build good models for these kinds of datasets, which can identify all classes accurately. This paper looks at what’s happening in this field lately and proposes a new way to understand the different techniques being used. They also review existing methods and compare them to imbalance learning approaches. Overall, this work gives an overview of the latest advancements in long-tailed visual learning. |
Keywords
* Artificial intelligence * Loss function
