AI Paper Atlas

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

Fourier Self-Supervision for Fine-Grained Generalized Category Discovery

Sarah Rastegar, Mina Ghadimi Atigh, Pascal Mettes, Yuki Asano, Cees Snoek

ECCV 2026 · Label-efficient and robust learning › Recognition and long-tail · approach: Contrastive representation theory · 0 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 low-frequency contrastive loss (ablation)SelEx (last 3 blocks) · CUB-200 (DINOv1)accuracy (All): 76.4 -> 77.5Improved
add high-frequency contrastive loss (ablation)SelEx (last 3 blocks) · CUB-200 (DINOv1)accuracy (All): 76.4 -> 78.8Improved
add frequency classification loss (ablation)SelEx (last 3 blocks) · CUB-200 (DINOv1)accuracy (All): 76.4 -> 78.0Improved
combine low and high contrastive losses (ablation)SelEx (last 3 blocks) · CUB-200 (DINOv1)accuracy (All): 76.4 -> 80.0Improved
add classification loss on top of both contrastive losses (ablation)FourEx · CUB-200 (DINOv1)accuracy (All): 80.0 -> 80.2Improved
add classification loss alone hurts known classes (ablation)SelEx (last 3 blocks) · CUB-200 (DINOv1)accuracy (Known): 72.4 -> 71.2Hurt
increase alpha_low to 1 from 0.1 (ablation)FourEx · CUB-200 (DINOv1)accuracy (All): 80.2 -> 69.2Hurt
use alpha_low 0.1 versus 0 (ablation)FourEx · CUB-200 (DINOv1)accuracy (All): 77.9 -> 80.2Improved
use alpha_high 0.1 versus 0 (ablation)FourEx · CUB-200 (DINOv1)accuracy (All): 78.1 -> 80.2Improved
use alpha_cls 0.1 versus 0 (ablation)FourEx · CUB-200 (DINOv1)accuracy (All): 79.7 -> 80.2Improved
raise SNR cutoff from 15 dB to 25 dB (ablation)FourEx · CUB-200 (DINOv1)accuracy (All): 80.2 -> 77.1Hurt
apply Fourier self-supervision to SelEx with DINOv1SelEx (last 3 blocks) · Stanford-Cars (DINOv2)accuracy (All): 82.2 -> 80.1Hurt

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.

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