Dissecting Generalized Category Discovery: Multiplex Consensus under Self-Deconstruction
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866912537722748928 |
|---|---|
| author | Tang, Luyao Huang, Kunze Chen, Chaoqi Yuan, Yuxuan Li, Chenxin Tu, Xiaotong Ding, Xinghao Huang, Yue |
| author_facet | Tang, Luyao Huang, Kunze Chen, Chaoqi Yuan, Yuxuan Li, Chenxin Tu, Xiaotong Ding, Xinghao Huang, Yue |
| contents | Human perceptual systems excel at inducing and recognizing objects across both known and novel categories, a capability far beyond current machine learning frameworks. While generalized category discovery (GCD) aims to bridge this gap, existing methods predominantly focus on optimizing objective functions. We present an orthogonal solution, inspired by the human cognitive process for novel object understanding: decomposing objects into visual primitives and establishing cross-knowledge comparisons. We propose ConGCD, which establishes primitive-oriented representations through high-level semantic reconstruction, binding intra-class shared attributes via deconstruction. Mirroring human preference diversity in visual processing, where distinct individuals leverage dominant or contextual cues, we implement dominant and contextual consensus units to capture class-discriminative patterns and inherent distributional invariants, respectively. A consensus scheduler dynamically optimizes activation pathways, with final predictions emerging through multiplex consensus integration. Extensive evaluations across coarse- and fine-grained benchmarks demonstrate ConGCD's effectiveness as a consensus-aware paradigm. Code is available at github.com/lytang63/ConGCD. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2508_10731 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Dissecting Generalized Category Discovery: Multiplex Consensus under Self-Deconstruction Tang, Luyao Huang, Kunze Chen, Chaoqi Yuan, Yuxuan Li, Chenxin Tu, Xiaotong Ding, Xinghao Huang, Yue Computer Vision and Pattern Recognition Machine Learning Human perceptual systems excel at inducing and recognizing objects across both known and novel categories, a capability far beyond current machine learning frameworks. While generalized category discovery (GCD) aims to bridge this gap, existing methods predominantly focus on optimizing objective functions. We present an orthogonal solution, inspired by the human cognitive process for novel object understanding: decomposing objects into visual primitives and establishing cross-knowledge comparisons. We propose ConGCD, which establishes primitive-oriented representations through high-level semantic reconstruction, binding intra-class shared attributes via deconstruction. Mirroring human preference diversity in visual processing, where distinct individuals leverage dominant or contextual cues, we implement dominant and contextual consensus units to capture class-discriminative patterns and inherent distributional invariants, respectively. A consensus scheduler dynamically optimizes activation pathways, with final predictions emerging through multiplex consensus integration. Extensive evaluations across coarse- and fine-grained benchmarks demonstrate ConGCD's effectiveness as a consensus-aware paradigm. Code is available at github.com/lytang63/ConGCD. |
| title | Dissecting Generalized Category Discovery: Multiplex Consensus under Self-Deconstruction |
| topic | Computer Vision and Pattern Recognition Machine Learning |
| url | https://arxiv.org/abs/2508.10731 |