AI Paper Atlas

Research map / Benchmarks / ImageNet-1K 1% labels

ImageNet-1K 1% labels leaderboard

12 papers at ICML, NeurIPS, ICLR, CVPR, AAAI, ICCV and ECCV, and the journals TPAMI, IJCV and JMLR report results on ImageNet-1K 1% labels. Best accuracy: 81.9 (Semi-ViM-Base, ICCV 2025).

Numbers are as each paper reports them, read by AI from its results tables and not checked by hand. Papers can differ in setting (training data, model size, extra data); the topic column shows what each paper is about. Open the paper before you compare or quote a number.

ImageNet-1K 1% labels · accuracy

12 papers, each with its best reported number for its own method; higher is better.

#accuracy ↑MethodPaperTopic
181.9Semi-ViM-Base
ViM-Base
Semi-ViM: Bidirectional State Space Model for Mitigating Label Imbalance in Semi-Supervised Learning
ICCV 2025 · Table 2
Semi-supervised learning
281.8Ours
ResNet-50
Train a One-Million-Way Instance Classifier for Unsupervised Visual Representation Learning
AAAI 2021 · Semi-supervised learning compa
Recognition and long-tail
377.4REACT (Locked-Text Gated-Image)
CLIP ViT-B/16
Learning Customized Visual Models With Retrieval-Augmented Knowledge
CVPR 2023 · Table: low-shot ImageNet-1K
CLIP and zero-shot
477.3I-JEPA
ViT-H/16 at 448 resolution
Self-Supervised Learning From Images With a Joint-Embedding Predictive Architecture
CVPR 2023 · Table: ImageNet-1% semi-superv
Self-supervised and contrastive
575.7MSN
ViT-B/4
Masked Siamese Networks for Label-Efficient Learning
ECCV 2022 · Table 2
Self-supervised and contrastive
667.5ShrinkMatch
ResNet-50
Shrinking Class Space for Enhanced Certainty in Semi-Supervised Learning
ICCV 2023 · Table 5
Semi-supervised learning
762.2USL-CLIP
FixMatch
Unsupervised Selective Labeling for More Effective Semi-Supervised Learning
ECCV 2022 · Table 6
Semi-supervised learning
856.2PMP (Barlow Twins)
ResNet-50 (Barlow Twins)
Exploring the Gap between Collapsed & Whitened Features in Self-Supervised Learning
ICML 2022 · Table 1
Self-supervised and contrastive
955.9DINO+SERE
ViT-S/16
SERE: Exploring Feature Self-Relation for Self-Supervised Transformer
TPAMI 2023 · Table 5
Self-supervised and contrastive
1055.8I-VNE+
ResNet-50
VNE: An Effective Method for Improving Deep Representation by Manipulating Eigenvalue Distribution
CVPR 2023 · Table: SSL semi-supervised eva
Self-supervised and contrastive
2 more papers on this leaderboard, the baselines reported on it and a CSV download are free with a scime account.
Sign in free to see all

Not affiliated with any conference, journal or the benchmark's authors. Corrections: contact@latenticlabs.com.