AI Paper Atlas

Research map / Benchmarks / PASCAL VOC 2007

PASCAL VOC 2007 leaderboard

194 papers at ICML, NeurIPS, ICLR, CVPR, AAAI, ICCV and ECCV, and the journals TPAMI, IJCV and JMLR report results on PASCAL VOC 2007. Best mAP: 97.1 (PatchCT, ICCV 2023).

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.

PASCAL VOC 2007 · mAP

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

#mAP ↑MethodPaperTopic
197.1PatchCT
ViT-B/16 + BERT
PatchCT: Aligning Patch Set and Label Set with Conditional Transport for Multi-Label Image Classification
ICCV 2023 · Table 3
Recognition and long-tail
296.2PAT-T + TResL
TResNetL
Counterfactual Reasoning for Multi-Label Image Classification via Patching-Based Training
ICML 2024 · Table: Comparison results on V
Recognition and long-tail
396ADD-GCN
ResNet-101
Attention-Driven Dynamic Graph Convolutional Network for Multi-Label Image Recognition
ECCV 2020 · Table 2
Recognition and long-tail
495.8CCD-R101
ResNet101
Contextual Debiasing for Visual Recognition With Causal Mechanisms
CVPR 2022 · Table 2
Recognition and long-tail
595.2ResNet-cut + CSRAResidual Attention: A Simple but Effective Method for Multi-Label Recognition
ICCV 2021 · Table 2
Recognition and long-tail
695MambaMLMambaML: Exploring State Space Models for Multi-Label Image Classification
ICCV 2025 · Table 3
Recognition and long-tail
795SSGRL (Ours, COCO pre-trained)
ResNet-101
Learning Semantic-Specific Graph Representation for Multi-Label Image Recognition
ICCV 2019 · Table 2
Recognition and long-tail
895WILDCAT
ResNet-101
WILDCAT: Weakly Supervised Learning of Deep ConvNets for Image Classification, Pointwise Localization and Segmentation
CVPR 2017 · Table 2
Weakly supervised detection and segmentation
994.4DSDL
ResNet-101
Deep Semantic Dictionary Learning for Multi-label Image Classification
AAAI 2021 · Table 1
Recognition and long-tail
1094ML-GCN (Re-weighted)
ResNet-101
Multi-Label Image Recognition With Graph Convolutional Networks
CVPR 2019 · Table 2
Knowledge and temporal graphs
155 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

PASCAL VOC 2007 · AP

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

#AP ↑MethodPaperTopic
176.5Cascade Eff-B7 NAS-FPN with Copy-Paste pre-training
Cascade Eff-B7 NAS-FPN
Simple Copy-Paste Is a Strong Data Augmentation Method for Instance Segmentation
CVPR 2021 · Table: PASCAL VOC detection re
Open-set and few-shot detection
263.9Cascade R-CNN + Rank-NMS
ResNet-101-FPN
Learning to Rank Proposals for Object Detection
ICCV 2019 · Table 5
CNN features
361.1Cascade D-PRD
ResNet101-FPN
Deformable Part Region Learning for Object Detection
AAAI 2022 · Table 5
CNN features
456.9DiLo-BASNet
MoCo / ResNet50-C4
Distilling Localization for Self-Supervised Representation Learning
AAAI 2021 · Table: Transfer learning for o
Open-set and few-shot detection
556.8RetinaNet w/ SWN
ResNeXt-101
Learning a Unified Sample Weighting Network for Object Detection
CVPR 2020 · Table 2
CNN features
655.3Grid R-CNN w FPN
ResNet-50
Grid R-CNN
CVPR 2019 · Table 4
CNN features
754.18SMVTEPSemi-Supervised Multi-View Multi-Label Learning with View-Specific Transformer and Enhanced Pseudo-Label
AAAI 2025 · Table 1
Semi-supervised learning
853.1AP-Loss
RetinaNet-ResNet50
AP-Loss for Accurate One-Stage Object Detection
TPAMI 2020 · Table 2
CNN features
953.1RetinaNet + AP-Loss
RetinaNet
Towards Accurate One-Stage Object Detection With AP-Loss
CVPR 2019 · Table: Comparison through diff
CNN features
1052.5WS-JDS FRCNN
Fast-RCNN
Cyclic Guidance for Weakly Supervised Joint Detection and Segmentation
CVPR 2019 · Table 2
Weakly supervised detection and segmentation
6 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

PASCAL VOC 2007 · AP50

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

#AP50 ↑MethodPaperTopic
186.5AutoFocus (1250^2 pixels)
ResNet-101
AutoFocus: Efficient Multi-Scale Inference
ICCV 2019 · Table 2
CNN features
284.1SPNetVOC
ResNet18
SP-NAS: Serial-to-Parallel Backbone Search for Object Detection
CVPR 2020 · Table 5
CNN features
383.9RetinaNet500 + AP-Loss
RetinaNet500, ResNet-101
Towards Accurate One-Stage Object Detection With AP-Loss
CVPR 2019 · Table: Detection results on VO
CNN features
483.6Faster R-CNN with RaMBO
ResNeXt101 32x4d
Optimizing Rank-Based Metrics With Blackbox Differentiation
CVPR 2020 · Table (object detection)
CNN features
582.5Ours
Faster R-CNN with FPN, ResNet-50
Mind the Gap: Polishing Pseudo Labels for Accurate Semi-supervised Object Detection
AAAI 2023 · Table 1
Weakly supervised detection and segmentation
682.3AP-Loss
RetinaNet-ResNet50
AP-Loss for Accurate One-Stage Object Detection
TPAMI 2020 · Table 2
CNN features
781.5De-biased Teacher
Faster R-CNN ResNet-50-FPN
De-biased Teacher: Rethinking IoU Matching for Semi-supervised Object Detection
AAAI 2023 · Table 3
Weakly supervised detection and segmentation
880.6SED
Faster R-CNN
Scale-Equivalent Distillation for Semi-Supervised Object Detection
CVPR 2022 · Table 1
Weakly supervised detection and segmentation
980.4ARSL
FCOS
Ambiguity-Resistant Semi-Supervised Learning for Dense Object Detection
CVPR 2023 · Table: Experimental results on
Weakly supervised detection and segmentation
1078.3MetricOpt + AP loss
Faster R-CNN, ResNet-50
MetricOpt: Learning To Optimize Black-Box Evaluation Metrics
CVPR 2021 · Table 5
Recognition and long-tail
3 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

PASCAL VOC 2007 · CorLoc

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

#CorLoc ↑MethodPaperTopic
183Yes/No human verification
AlexNet
We Don't Need No Bounding-Boxes: Training Object Class Detectors Using Only Human Verification
CVPR 2016 · text
Weakly supervised detection and segmentation
280.6WSOVOD
RN18-WS-MRRP
Weakly Supervised Open-Vocabulary Object Detection
AAAI 2024 · Table: Ablation study of DAFE
Weakly supervised detection and segmentation
380.3CaT_5 (COCO-65)
VGG16 Faster R-CNN
CaT: Weakly Supervised Object Detection With Category Transfer
ICCV 2021 · Table 1
Weakly supervised detection and segmentation
469.4Proposed
Faster R-CNN
What Can Human Sketches Do for Object Detection?
CVPR 2023 · Table: category-level object d
Weakly supervised detection and segmentation
566.5OICR + W-RPN
VGG16
You Reap What You Sow: Using Videos to Generate High Precision Object Proposals for Weakly-Supervised Object Detection
CVPR 2019 · Table 2
Weakly supervised detection and segmentation
663.8WSR18
WS-ResNet
UWSOD: Toward Fully-Supervised-Level Capacity Weakly Supervised Object Detection
NeurIPS 2020 · Table 2
Weakly supervised detection and segmentation
763.1WSDDN + all four principles (ResNet18-WS)
ResNet18
Enabling Deep Residual Networks for Weakly Supervised Object Detection
ECCV 2020 · Table 4
Weakly supervised detection and segmentation
862.2ICE
DeiT-S
Adversarial Normalization: I Can Visualize Everything (ICE)
CVPR 2023 · Table 4
Attention and transformers
960.8+TST (SEM) + ILSVRC-dets
Deep MIL + objectness + multifolding
Weakly Supervised Object Localization Using Things and Stuff Transfer
ICCV 2017 · Table 4
Weakly supervised detection and segmentation
1055.4DINOv2+reg
ViT-L
Vision Transformers Need Registers
ICLR 2024 · Table: Unsupervised Object Dis
Attention and transformers
1 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.