AI Paper Atlas

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

Parametric Information Maximization for Generalized Category Discovery

Florent Chiaroni, Jose Dolz, Ziko Imtiaz Masud, Amar Mitiche, Ismail Ben Ayed

ICCV 2023 · Label-efficient and robust learning › Recognition and long-tail · approach: Contrastive representation theory · 51 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
compute marginal entropy over labeled plus unlabeled data instead of unlabeled only (ablation)PIM · CUB (fine-grained)ACC All: 58.3 -> 62.7Improved
add supervision constraint (CE on labeled samples) (ablation)PIM · CUB (fine-grained)ACC All: 56.6 -> 62.7Improved
add marginal entropy term to conditional entropy objective (ablation)PIM · CUB (fine-grained)ACC All: 38.6 -> 62.7Improved
minimize conditional entropy only (no marginal entropy, no constraint) (ablation)PIM · CUB (fine-grained)ACC All: 6.1 vs 62.7Hurt
learn lambda via bi-level optimization instead of fixed standard MI (ablation)PIM · Herbarium19 (long-tailed) / CUBACC AllImproved
PIM versus standard InfoMax baselines TIM on fine-grained CUBTIM · CUB (fine-grained)ACC All: 53.4 -> 62.7Improved
PIM versus ORCA ViTB16 on simple CIFAR-10PIM · CIFAR10ACC All: 94.7 vs 97.1Hurt
use PIM instead of k-means for estimating number of classesMax-ACC class-number search · fine-grained datasets mean errorErr: 18% -> 15%Improved
PIM versus Max-ACC (GCD) on CIFAR-100 class-number estimationMax-ACC · CIFAR100 K estimationErr: 0% vs 5%Hurt

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