Official repository for GCR rerank, a GCN-based reranking method for both image and video re-ID

Last update: Nov 22, 2022

Related tags

Deep Learning GCN_rerank

Overview

[ICASSP] Graph Convolution for Re-ranking in Person Re-identification

The official repository for GCR rerank, a GCN-based reranking method for both image and video re-ID.

Environment

We use python 3.7/torch 1.6/torchvision 0.7.0.

Extracted features

We provide Market1501/MARS features from reid-strong-baseline at Google Drive.

Command Lines

Run GCRV rerank with basic settings on Market1501

python eval_rerank.py --config_file=config/market.yml

Run PVG only

python eval_rerank.py --config_file=config/market.yml PVG.ENABLE_PVG True GCR.ENABLE_GCR False

Run GCR only

python eval_rerank.py --config_file=config/market.yml PVG.ENABLE_PVG False GCR.ENABLE_GCR True

RUN GCRV on video reid dataset(MARS)

python eval_rerank.py --config_file=config/mars.yml

Run other rerank methods: (baseline, k_reciprocal, ecn, ecn_orig, lbr, qe)

python eval_rerank.py --config_file=config/market.yml COMMON.RERANK_TYPE baseline

Thanks

State-of-the-art reranking method inlucidng K_reciprocal, ECN, LBR

Citation

If you find our work useful in your research, please consider citing:

@inproceedings{zhang2022graph,
 title={Graph Convolution for Re-ranking in Person Re-identification},
 author={Zhang, Yuqi and Qian Qi and Liu Chong and Chen, Weihua and Wang Fan and Li Hao and Jin Rong},
 journal={arXiv preprint arXiv:2107.02220},
 year={2022}
}

Official repository for GCR rerank, a GCN-based reranking method for both image and video re-ID

Related tags

Overview

[ICASSP] Graph Convolution for Re-ranking in Person Re-identification

Environment

Extracted features

Command Lines

Thanks

Citation

Owner

Parallel Latent Tree-Induction for Faster Sequence Encoding

Distinguishing Commercial from Editorial Content in News

Generative Modelling of BRDF Textures from Flash Images [SIGGRAPH Asia, 2021]

Efficient electromagnetic solver based on rigorous coupled-wave analysis for 3D and 2D multi-layered structures with in-plane periodicity

Neural Cellular Automata + CLIP

本步态识别系统主要基于GaitSet模型进行实现

Code for ICCV 2021 paper Graph-to-3D: End-to-End Generation and Manipulation of 3D Scenes using Scene Graphs

This repo holds the code of TransFuse: Fusing Transformers and CNNs for Medical Image Segmentation

A no-BS, dead-simple training visualizer for tf-keras

Simple image captioning model - CLIP prefix captioning.

《Improving Unsupervised Image Clustering With Robust Learning》(2020)

Official PyTorch code for the paper: "Point-Based Modeling of Human Clothing" (ICCV 2021)

Code and results accompanying our paper titled Mixture Proportion Estimation and PU Learning: A Modern Approach at Neurips 2021 (Spotlight)

Joint parameterization and fitting of stroke clusters

Simple, but essential Bayesian optimization package

Efficient semidefinite bounds for multi-label discrete graphical models.

QMagFace: Simple and Accurate Quality-Aware Face Recognition

PointCloud Annotation Tools, support to label object bound box, ground, lane and kerb

Official implementation of the paper ``Unifying Nonlocal Blocks for Neural Networks'' (ICCV'21)

A state of the art of new lightweight YOLO model implemented by TensorFlow 2.