Person Re Identification On Cuhk03

Metrics

Rank-1

Rank-10

Rank-5

Results

Performance results of various models on this benchmark

				Paper Title
AlignedReID (RK)	97.8	99.8	99.6	AlignedReID: Surpassing Human-Level Performance in Person Re-Identification
Deep Constrained Dominant Sets	95.8	-	99.1	Deep Constrained Dominant Sets for Person Re-identification
FD-GAN	92.6	-	-	FD-GAN: Pose-guided Feature Distilling GAN for Robust Person Re-identification
ProNet++ (ResNet50+RK)	90.6	-	-	Rethinking Person Re-identification from a Projection-on-Prototypes Perspective
TriNet	89.63	-	99.01	In Defense of the Triplet Loss for Person Re-Identification
ProNet++	85.2	-	-	Rethinking Person Re-identification from a Projection-on-Prototypes Perspective
Weakly Supervised Pre-training (ResNet50+BDB)	84.7	-	-	Large-Scale Pre-training for Person Re-identification with Noisy Labels
VI+LSRO 3	84.6	-	-	Unlabeled Samples Generated by GAN Improve the Person Re-identification Baseline in vitro
DLCE	83.4	-	-	A Discriminatively Learned CNN Embedding for Person Re-identification
Unsupervised Pre-training (ResNet50+BDB)	81.9	-	-	Unsupervised Pre-training for Person Re-identification
OIM Loss 45	77.5	-	-	Joint Detection and Identification Feature Learning for Person Search
DG-Net	65.6	-	-	Joint Discriminative and Generative Learning for Person Re-identification
TriNet + Era + Reranking (ACNet, bs=32)	64.8	-	-	Adaptively Connected Neural Networks
k-reciprocal 46	61.6	-	-	Re-ranking Person Re-identification with k-reciprocal Encoding
UTAL	56.3	-	-	Unsupervised Tracklet Person Re-Identification
HA-CNN	41.7	-	-	Harmonious Attention Network for Person Re-Identification
OSNet	-	-	-	Omni-Scale Feature Learning for Person Re-Identification
UniHCP (finetune)	-	-	-	UniHCP: A Unified Model for Human-Centric Perceptions
Proposed SGGNN	-	-	-	Person Re-identification with Deep Similarity-Guided Graph Neural Network

0 of 19 row(s) selected.

Person Re Identification On Cuhk03 | SOTA | HyperAI