AI Paper Atlas

Research map / Benchmarks / CUB-200

CUB-200 leaderboard

89 papers at ICML, NeurIPS, ICLR, CVPR, AAAI, ICCV and ECCV, and the journals TPAMI, IJCV and JMLR report results on CUB-200. Best accuracy: 99 (Deep Novelty, CVPR 2019).

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.

CUB-200 · accuracy

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

51 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

CUB-200 · R@1

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

#R@1 ↑MethodPaperTopic
188.5VPTSP-G (ViT-B16/512)
ViT-B16/512
Learning Semantic Proxies from Visual Prompts for Parameter-Efficient Fine-Tuning in Deep Metric Learning
ICLR 2024 · Table: comparison with state-o
CLIP and zero-shot
282.7IRGen (beam search)IRGen: Generative Modeling for Image Retrieval
ECCV 2024 · Table 2
Recognition and long-tail
378.69VADSL-PA
ResNet-50
Volume-Aware Distance for Robust Similarity Learning
ICML 2025 · Table 2
Point cloud classification
478.3LOCORE-baseLOCORE: Image Re-ranking with Long-Context Sequence Modeling
CVPR 2025 · Table 2
Autoregressive and latent generation
572.9PA+DADA(R50)
ResNet50
Towards Improved Proxy-Based Deep Metric Learning via Data-Augmented Domain Adaptation
AAAI 2024 · Table 1
Domain adaptation and generalization
671.4Multi-Similarity + Metrix
ResNet-50
It Takes Two to Tango: Mixup for Deep Metric Learning
ICLR 2022 · Table: Improving the SOTA with
Recognition and long-tail
771.3Bayes vMF
Densenet161
Bayesian Triplet Loss: Uncertainty Quantification in Image Retrieval
ICCV 2021 · Table 1
Uncertainty and Gaussian processes
870.1AutoLR
ResNet-50
AutoLR: Layer-wise Pruning and Auto-tuning of Learning Rates in Fine-tuning of Deep Networks
AAAI 2021 · Comparison with state-of-the-a
Pruning and efficient tuning
965.7MS
Inception-BN
Multi-Similarity Loss With General Pair Weighting for Deep Metric Learning
CVPR 2019 · Table: Recall@K performance on
Recognition and long-tail
1065.4DeML(I=3,J=3)
GooglenetV1
Hybrid-Attention Based Decoupled Metric Learning for Zero-Shot Image Retrieval
CVPR 2019 · Table: retrieval results
CLIP and zero-shot
8 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

CUB-200 · FID

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

#FID ↓MethodPaperTopic
12.13VQCTCodebook Transfer with Part-of-Speech for Vector-Quantized Image Modeling
CVPR 2024 · Table 1
Human and embodied modeling
23.08VQ-GAN+LG
VQ-GAN
LG-VQ: Language-Guided Codebook Learning
NeurIPS 2024 · Image reconstruction table
Human and embodied modeling
34.03TA-CVQ
CVQ
Towards Improved Text-Aligned Codebook Learning: Multi-Hierarchical Codebook-Text Alignment with Long Text
CVPR 2025 · Table 1
Human and embodied modeling
49.11HyperSphereDiff (vMF)Harmonizing Geometry and Uncertainty: Diffusion with Hyperspheres
ICML 2025 · Table: Performance comparison
Fast diffusion sampling
510.32VQ-Diffusion-FVector Quantized Diffusion Model for Text-to-Image Synthesis
CVPR 2022 · Table: FID comparison of diffe
Text-to-image diffusion
610.78LCP-GAN (S)
StyleGAN2-ADA
Exploring Intra-Class Variation Factors With Learnable Cluster Prompts for Semi-Supervised Image Synthesis
CVPR 2023 · Table 3
Semi-supervised learning
712.61Ours Sample-Inter (T=80)Discrete Contrastive Diffusion for Cross-Modal Music and Image Generation
ICLR 2023 · Table 2
Autoregressive and latent generation
815.19DAE-GANDAE-GAN: Dynamic Aspect-Aware GAN for Text-to-Image Synthesis
ICCV 2021 · Table 2
Text-to-image diffusion
915.4ReACGAN
BigGAN
Rebooting ACGAN: Auxiliary Classifier GANs with Stable Training
NeurIPS 2021 · Table 1
Recognition and long-tail
1015.85VisCoIN
ResNet50 + StyleGAN2-ADA
Restyling Unsupervised Concept Based Interpretable Networks with Generative Models
ICLR 2025 · Table 2a
Recognition and long-tail
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.