HyperAI초신경

홈 뉴스 최신 연구 논문 튜토리얼 데이터셋 백과사전 SOTA LLM 모델 GPU 랭킹 컨퍼런스

한국어

시스템 설정

HyperAI초신경

Text Based Person Retrieval On Icfg Pedes

평가 지표

R@1

R@10

R@5

mAP

평가 결과

이 벤치마크에서 각 모델의 성능 결과

모델 이름	R@1	R@10	R@5	mAP	Paper Title	Repository
RaSa	65.28	85.12	80.4	41.29	RaSa: Relation and Sensitivity Aware Representation Learning for Text-based Person Search	-
PLIP-RN50	64.25	86.32	80.88	-	PLIP: Language-Image Pre-training for Person Representation Learning
APTM	68.51	-	-	41.22	Towards Unified Text-based Person Retrieval: A Large-scale Multi-Attribute and Language Search Benchmark
Filtering-WoRA(Small)	68.35	87.53	83.10	42.60	From Data Deluge to Data Curation: A Filtering-WoRA Paradigm for Efficient Text-based Person Search	-
MARS	67.60	85.79	81.47	44.93	MARS: Paying more attention to visual attributes for text-based person search
VGSG (ViT-Base)	63.05	84.36	78.43	-	VGSG: Vision-Guided Semantic-Group Network for Text-based Person Search	-
IRRA	63.46	85.82	80.25	38.06	Cross-Modal Implicit Relation Reasoning and Aligning for Text-to-Image Person Retrieval	-
SRCF	57.18	-	-	-	A Simple and Robust Correlation Filtering Method for Text-based Person Search
TBPS-CLIP (ViT-B/16)	65.05	85.47	80.34	39.83	An Empirical Study of CLIP for Text-based Person Search
RDE	67.68	87.36	82.47	40.06	Noisy-Correspondence Learning for Text-to-Image Person Re-identification
SSAN	54.23	-	-	-	Semantically Self-Aligned Network for Text-to-Image Part-aware Person Re-identification
CADA	-	-	-	39.85	Cross-Modal Adaptive Dual Association for Text-to-Image Person Retrieval	-

0 of 12 row(s) selected.