HyperAI초신경
홈뉴스최신 연구 논문튜토리얼데이터셋백과사전SOTALLM 모델GPU 랭킹컨퍼런스
전체 검색
소개
HyperAI초신경

Pedestrian Attribute Recognition On Pa 100K

평가 지표

Accuracy

평가 결과

이 벤치마크에서 각 모델의 성능 결과

모델 이름
Accuracy
Paper TitleRepository
strongBaseline(ours)78.56Rethinking of Pedestrian Attribute Recognition: Realistic Datasets with Efficient Method
PATH (Partial FT)90.8HumanBench: Towards General Human-centric Perception with Projector Assisted Pretraining
HP-net72.19%HydraPlus-Net: Attentive Deep Features for Pedestrian Analysis
UniHCP (finetune)86.18UniHCP: A Unified Model for Human-Centric Perceptions
Label2Label79.23Label2Label: A Language Modeling Framework for Multi-Attribute Learning
APTM80.17Towards Unified Text-based Person Retrieval: A Large-scale Multi-Attribute and Language Search Benchmark
Hulk(Finetune, ViT-B)87.85Hulk: A Universal Knowledge Translator for Human-Centric Tasks
Hulk(Finetune, ViT-L)88.97Hulk: A Universal Knowledge Translator for Human-Centric Tasks
Attribute-Specific Localization77.08%Improving Pedestrian Attribute Recognition With Weakly-Supervised Multi-Scale Attribute-Specific Localization
C2T-Net87.2C2T-Net: Channel-Aware Cross-Fused Transformer-Style Networks for Pedestrian Attribute Recognition
SOLIDER86.38Beyond Appearance: a Semantic Controllable Self-Supervised Learning Framework for Human-Centric Visual Tasks
0 of 11 row(s) selected.