AI Paper Atlas

Research map / Benchmarks / PETA

PETA leaderboard

14 papers at ICML, NeurIPS, ICLR, CVPR, AAAI, ICCV and ECCV, and the journals TPAMI, IJCV and JMLR report results on PETA. Best mA: 92.25 (LLM-PAR, AAAI 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.

PETA · mA

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

#mA ↑MethodPaperTopic
192.25LLM-PAR
EVA-ViT-G; Vicuna-7B
Pedestrian Attribute Recognition: A New Benchmark Dataset and a Large Language Model Augmented Framework
AAAI 2025 · Table: Comparison with SOTA me
Human and embodied modeling
290.27JGEL
CLIP ViT-B/16
Joint Implicit and Explicit Language Learning for Pedestrian Attribute Recognition
AAAI 2026 · Table 1
CLIP and zero-shot
389.65EVSITP
CLIP
Enhanced Visual-Semantic Interaction with Tailored Prompts for Pedestrian Attribute Recognition
CVPR 2025 · Table 2
Cross-modal learning
488.78UniHCP (finetune)
ViT-B
UniHCP: A Unified Model for Human-Centric Perceptions
CVPR 2023 · Transfer results table
Open-set and few-shot detection
588.59FRDL + GOAT
ConvNeXt-base
Pedestrian Attribute Recognition as Label-balanced Multi-label Learning
ICML 2024 · Table 1
Recognition and long-tail
688.36HAP
ViT-Base
HAP: Structure-Aware Masked Image Modeling for Human-Centric Perception
NeurIPS 2023 · Table 1
Human pose estimation
787.1SOFAFormerSelective and Orthogonal Feature Activation for Pedestrian Attribute Recognition
AAAI 2024 · Table 1
Recognition and long-tail
887.07DAFL
ResNet50
Learning Disentangled Attribute Representations for Robust Pedestrian Attribute Recognition
AAAI 2022 · Table 2
Recognition and long-tail
986.96JLACRelation-Aware Pedestrian Attribute Recognition with Graph Convolutional Networks
AAAI 2020 · Table 2-4
Knowledge and temporal graphs
1086.52SSC_soft
ResNet50
Spatial and Semantic Consistency Regularizations for Pedestrian Attribute Recognition
ICCV 2021 · Table 1
Recognition and long-tail
4 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

PETA · F1

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

#F1 ↑MethodPaperTopic
190.39LLM-PAR
EVA-ViT-G; Vicuna-7B
Pedestrian Attribute Recognition: A New Benchmark Dataset and a Large Language Model Augmented Framework
AAAI 2025 · Table: Comparison with SOTA me
Human and embodied modeling
289.22JGEL
CLIP ViT-B/16
Joint Implicit and Explicit Language Learning for Pedestrian Attribute Recognition
AAAI 2026 · Table 1
CLIP and zero-shot
389.03FRDL + GOAT
ConvNeXt-base
Pedestrian Attribute Recognition as Label-balanced Multi-label Learning
ICML 2024 · Table 1
Recognition and long-tail
488.42Ours (visual-semantic graph reasoning)
ResNet-50
Visual-Semantic Graph Reasoning for Pedestrian Attribute Recognition
AAAI 2019 · Table 1
Knowledge and temporal graphs
587.45JLACRelation-Aware Pedestrian Attribute Recognition with Graph Convolutional Networks
AAAI 2020 · Table 2-4
Knowledge and temporal graphs
686.99SSC_soft
ResNet50
Spatial and Semantic Consistency Regularizations for Pedestrian Attribute Recognition
ICCV 2021 · Table 1
Recognition and long-tail
786.85Ours
BN-Inception
Improving Pedestrian Attribute Recognition With Weakly-Supervised Multi-Scale Attribute-Specific Localization
ICCV 2019 · Table 3
Weakly supervised detection and segmentation
886.7RCRecurrent Attention Model for Pedestrian Attribute Recognition
AAAI 2019 · Table 3
Attention and transformers
986.46OursDeep Imbalanced Attribute Classification using Visual Attention Aggregation
ECCV 2018 · Table 3
Recognition and long-tail
1086.4DAFL
ResNet50
Learning Disentangled Attribute Representations for Robust Pedestrian Attribute Recognition
AAAI 2022 · Table 2
Recognition and long-tail

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