深度自学笔记(Person Search)

一些乱七八糟给自己看的东西,并不觉得有人会看😅

Overview

行人搜索的工作主要是将行人检测(detection)以及行人重识别(Re-ID)整合在一起的工作。相比被广泛研究的单纯的行人重识别要更加贴近于现实的应用。

Paper Source Related Material
A Discriminatively Learned Feature Embedding Based on Multi-Loss Fusion For Person Search ICASSP 2018 -
Correlation Based Identity Filter: An Efficient Framework for Person Search ICIG2017 -
Joint Detection and Identification Feature Learning for Person Search CVPR2017 https://github.com/ShuangLI59/person_search
Enhanced Deep Feature Representation for Person Search CCCV2017 -
Fusion-Attention Network for person search with free-form natural language Pattern Recognition Letters 2018 -
IAN: The Individual Aggregation Network for Person Search Pattern Recognition 2019 -
Instance Enhancing Loss: Deep Identity-Sensitive Feature Embedding For Person Search ICIP2018 -
Neural Person Search Machines ICCV2017 -
Person Re-identification in the Wild CVPR2017 https://github.com/liangzheng06/PRW-baseline
Person Search by Multi-Scale Matching ECCV2018 -
Person Search in Videos with One Portrait Through Visual and Temporal Links ECCV2018 http://qqhuang.cn/projects/eccv18-person-search/ https://github.com/hqqasw/person-search-PPCC
Person Search via A Mask-Guided Two-Stream CNN Model ECCV 2018 -
Person Search with Natural Language Description CVPR2017 https://github.com/ShuangLI59/Person-Search-with-Natural-Language-Description
Transferring a Semantic Representation for Person Re-Identification and Search CVPR2015 -
Person Search in a Scene by Jointly Modeling People Commonness and Person Uniqueness ACM MM 2014 -
RCAA: Relational context-aware agents for person search ECCV2018 -
End-to-End Detection and Re-identification Integrated Net for Person Search ACCV2018 -
Learning Context Graph for Person Search CVPR2019 Oral https://github.com/sjtuzq/person_search_gcn
Query-guided End-to-End Person Search CVPR2019 https://github.com/munjalbharti/Query-guided-End-to-End-Person-Search
Partially Separated Networks for Person Search PCM2018 -
A cascaded multitask network with deformable spatial transform on person search International Journal of Advanced Robotic Systems -
Multilevel Collaborative Attention Network for Person Search ACCV2018 -
Enhancing Person Retrieval with Joint Person Detection, Attribute Learning, and Identification PCM2018 -
Spatial Invariant Person Search Network PRCV2018 https://github.com/liliangqi/person_search
FMT: fusing multi-task convolutional neural network for person search Multimedia Tools and Applications -
Segmentation Mask Guided End-to-End Person Search arXiv -
Dhff: Robust Multi-Scale Person Search by Dynamic Hierarchical Feature Fusion ICIP 2019 -
Comprehensive Samples Constrain for Person Search VCIP 2018 -
Fast Person Search Pipeline ICME 2019 -
End-To-End Person Search Sequentially Trained On Aggregated Dataset Image Processing -
Scale Voting With Pyramidal Feature Fusion Network for Person Search IEEE Access -
Person Search System Using Clothing Features Electronics and Communications in Japan -
Structure-aware person search with self-attention and online instance aggregation matching Neurocomputing https://github.com/gggcy/person_search
Knowledge Distillation for End-to-End Person Search arXiv -
Re-ID Driven Localization Refinement for Person Search ICCV 2019 -

Paper of Interest List

Paper Source Related Material
Attribute-based Person Retrieval and Search in Video Sequences AVSS2018 -
Fast Open-World Person Re-Identification Image Processing -
Re-ID done right: towards good practices for person re-identification arXiv -
Weakly Supervised Person Re-Identification CVPR2019 -
Multimodal clothing recognition for semantic search in unconstrained surveillance imagery VCIP -
Fusion-Attention Network for person search with free-form natural language PRL -
Cascade Attention Network for Person Search: Both Image and Text-Image Similarity Selection arXiv -
Deep Adversarial Graph Attention Convolution Network for Text-Based Person Search ACM MM -
Language Person Search with Mutually Connected Classification Loss ICASSP 2019 -
Improving Text-Based Person Search by Spatial Matching and Adaptive Threshold WACV 2018 -

Dataset

该领域主要只有两个数据集 PRW 以及 CUHK-SYSU

Docker Image for JDI-PS

做了docker镜像:https://hub.docker.com/r/4f5da2/person_search

Reminders for the Docker Image

  • tools/demo.py中的import matplotlib下添加matplotlib.use('Agg')来避免对GUI相关功能的调用(因为是在docker里)

  • 由于protobuf版本发生变化,需在lib/fast_rcnn/train.py中增加一行import google.protobuf.text_format

  • 按照Github上对应的readme进行编译,填入docker镜像中cuDNN的路径如下:cmake .. -DUSE_MPI=ON -DCUDNN_INCLUDE=/usr/include -DCUDNN_LIBRARY=/usr/lib/x86_64-linux-gnu/libcudnn.so