Summary of Neighborhood Commonality-aware Evolution Network For Continuous Generalized Category Discovery, by Ye Wang et al.
Neighborhood Commonality-aware Evolution Network for Continuous Generalized Category Discovery
by Ye Wang, Yaxiong Wang, Guoshuai Zhao, Xueming Qian
First submitted to arxiv on: 7 Dec 2024
Categories
- Main: Computer Vision and Pattern Recognition (cs.CV)
- Secondary: Artificial Intelligence (cs.AI)
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 proposed Neighborhood Commonality-aware Evolution Network (NCENet) is a novel learning framework that continually discovers novel classes from unlabelled image sets while maintaining performance on old classes. The NCENet consists of two main components: the Neighborhood Commonality-aware Representation Learning (NCRL), which learns discriminative representations for novel classes by exploiting local commonalities derived neighborhoods, and the Bi-level Contrastive Knowledge Distillation (BCKD) module, which leverages contrastive learning to perceive the learning and learned knowledge and conducts knowledge distillation. The NCENet is evaluated on CIFAR10, CIFAR100, and Tiny-ImageNet datasets, demonstrating superior performance compared to previous state-of-the-art methods. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary NCENet is a new way to learn from pictures without labels. It’s like teaching someone to recognize new objects by showing them similar ones. The model has two parts: one that learns to tell apart new and old objects, and another that helps it remember what it learned before. This makes NCENet good at recognizing both old and new objects. In tests, NCENet did better than other models on some pictures. |
Keywords
» Artificial intelligence » Knowledge distillation » Representation learning
