AI Paper Atlas

Research map / Label-efficient and robust learning / Recognition and long-tail

ProtoGCD: Unified and Unbiased Prototype Learning for Generalized Category Discovery

Shijie Ma, Fei Zhu, Xu-Yao Zhang, Cheng‐Lin Liu

TPAMI 2025 · Label-efficient and robust learning › Recognition and long-tail · approach: Contrastive representation theory · 32 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add DAPL pseudo-labeling to contrastive learning (ablation)ProtoGCD · CIFAR100All accuracy: 64.0 -> 65.8Improved
add entropy regularization without DAPL (ablation)ProtoGCD · CIFAR100All accuracy: 64.0 -> 30.1Hurt
add entropy regularization on top of DAPL (ablation)ProtoGCD · CIFAR100All accuracy: 65.8 -> 79.7Improved
add entropy regularization (with separation) (ablation)ProtoGCD · FGVC-AircraftAll accuracy: 38.0 -> 56.8Improved
add inter-class separation regularization (ablation)ProtoGCD · CIFAR100All accuracy: 79.7 -> 81.9Improved
add inter-class separation regularization (ablation)ProtoGCD · FGVC-AircraftAll accuracy: 54.4 -> 56.8Improved
add marginal entropy maximization on long-tailed Herbarium19 (ablation)ProtoGCD · Herbarium19All accuracy: 29.4 -> 44.5Improved
add small entropy weight 0.1 on Herbarium19 (ablation)ProtoGCD · Herbarium19All accuracy: 29.4 -> 36.2Improved
add sharpening of soft pseudo-labels (ablation)ProtoGCD DAPL · CIFAR100 and AircraftAll accuracy: ~20%Improved
use adaptive hard/soft ratio instead of fixed 1:1 (ablation)ProtoGCD DAPL · CIFAR100 and AircraftAll accuracy: ~2% lower with fixed 1:1Improved
vary ramp-up epochs e_ramp, optimum near 125 (ablation)ProtoGCD · CIFAR100 and AircraftAll accuracy: stable within [75,150]No clear effect
use DINOv2 initialization instead of DINOProtoGCD · CUBAll accuracy: 63.2 -> 74.9Improved

Later papers that compared with this method

3 later accepted papers report ProtoGCD as a baseline in their results tables; 3 of them report a better number on most of the results they share with it.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy