Summary of Drupi: Dataset Reduction Using Privileged Information, by Shaobo Wang et al.
DRUPI: Dataset Reduction Using Privileged Information
by Shaobo Wang, Yantai Yang, Shuaiyu Zhang, Chenghao Sun, Weiya Li, Xuming Hu, Linfeng Zhang
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 Dataset Reduction (DR) aims to condense large datasets into smaller subsets while maintaining performance on target tasks. Existing approaches mainly focus on pruning or synthesizing data in the same format as the original dataset, typically input data and corresponding labels. In this paper, we introduce Dataset Reduction Using Privileged Information (DRUPI), which enriches DR by synthesizing privileged information alongside the reduced dataset. This privileged information can take the form of feature labels or attention labels, providing auxiliary supervision to improve model learning. Our findings show that effective feature labels must strike a balance between being overly discriminative and excessively diverse, with a moderate level proving optimal for improving the reduced dataset’s efficacy. Extensive experiments on ImageNet, CIFAR-10/100, and Tiny ImageNet demonstrate that DRUPI integrates seamlessly with existing dataset reduction methods, offering significant performance gains. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper is about making big datasets smaller while keeping their helpfulness. Right now, people are trying to do this by getting rid of some data or creating new fake data. But what if we could create even more helpful information along with the reduced dataset? This would give machines a better chance to learn from it. The researchers in this paper came up with a new way to do just that, which they call Dataset Reduction Using Privileged Information (DRUPI). They found that making the right kind of extra information can make a big difference in how well the machine learns. They tested this idea on several different datasets and showed that it works really well. |
Keywords
» Artificial intelligence » Attention » Pruning
