# Weakly Supervised Person Search with Region Siamese Networks

arXiv: https://arxiv.org/abs/2109.06109 · HTML: https://mlsanity.com/paper/2109.06109/

## Key facts

- Cited by 14 papers in the map (data through 14 Oct 2026).
- Builds on 29 of its 47 references that are in the map.
- First posted on arXiv 13 Sep 2021.

## Builds on (most cited first, 29 of 29)

- [Deep Residual Learning for Image Recognition](https://mlsanity.com/paper/1512.03385/) (2015)
- [A Simple Framework for Contrastive Learning of Visual Representations](https://mlsanity.com/paper/2002.05709/) (2020)
- [Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks](https://mlsanity.com/paper/1506.01497/) (2015)
- [Momentum Contrast for Unsupervised Visual Representation Learning](https://mlsanity.com/paper/1911.05722/) (2019)
- [Feature Pyramid Networks for Object Detection](https://mlsanity.com/paper/1612.03144/) (2016)
- [Exploring Simple Siamese Representation Learning](https://mlsanity.com/paper/2011.10566/) (2020)
- [Unsupervised Feature Learning via Non-Parametric Instance-level Discrimination](https://mlsanity.com/paper/1805.01978/) (2018)
- [MMDetection: Open MMLab Detection Toolbox and Benchmark](https://mlsanity.com/paper/1906.07155/) (2019)
- [Learning deep representations by mutual information estimation and maximization](https://mlsanity.com/paper/1808.06670/) (2018)
- [Fully-Convolutional Siamese Networks for Object Tracking](https://mlsanity.com/paper/1606.09549/) (2016)
- [Circle Loss: A Unified Perspective of Pair Similarity Optimization](https://mlsanity.com/paper/2002.10857/) (2020)
- [RepPoints: Point Set Representation for Object Detection](https://mlsanity.com/paper/1904.11490/) (2019)
- [Unsupervised Embedding Learning via Invariant and Spreading Instance Feature](https://mlsanity.com/paper/1904.03436/) (2019)
- [Self-paced Contrastive Learning with Hybrid Memory for Domain Adaptive Object Re-ID](https://mlsanity.com/paper/2006.02713/) (2020)
- [Joint Detection and Identification Feature Learning for Person Search](https://mlsanity.com/paper/1604.01850/) (2016)
- [Person Re-identification in the Wild](https://mlsanity.com/paper/1604.02531/) (2016)
- [Efficient Parameter-free Clustering Using First Neighbor Relations](https://mlsanity.com/paper/1902.11266/) (2019)
- [Hierarchical Clustering with Hard-batch Triplet Loss for Person Re-identification](https://mlsanity.com/paper/1910.12278/) (2019)
- [Learning Context Graph for Person Search](https://mlsanity.com/paper/1904.01830/) (2019)
- [Person Search via A Mask-Guided Two-Stream CNN Model](https://mlsanity.com/paper/1807.08107/) (2018)
- [Re-ID Driven Localization Refinement for Person Search](https://mlsanity.com/paper/1909.08580/) (2019)
- [Query-guided End-to-End Person Search](https://mlsanity.com/paper/1905.01203/) (2019)
- [Person Search by Multi-Scale Matching](https://mlsanity.com/paper/1807.08582/) (2018)
- [Neural Person Search Machines](https://mlsanity.com/paper/1707.06777/) (2017)
- [Robust Partial Matching for Person Search in the Wild](https://mlsanity.com/paper/2004.09329/) (2020)
- [IAN: The Individual Aggregation Network for Person Search](https://mlsanity.com/paper/1705.05552/) (2017)
- [Decoupled and Memory-Reinforced Networks: Towards Effective Feature Learning for One-Step Person Search](https://mlsanity.com/paper/2102.10795/) (2021)
- [Diverse Knowledge Distillation for End-to-End Person Search](https://mlsanity.com/paper/2012.11187/) (2020)
- [Tasks Integrated Networks: Joint Detection and Retrieval for Image Search](https://mlsanity.com/paper/2009.01438/) (2020)

## Built on by (most cited first, 14 of 14)

- [Domain Adaptive Person Search](https://mlsanity.com/paper/2207.11898/) (2022)
- [Self-similarity Driven Scale-invariant Learning for Weakly Supervised Person Search](https://mlsanity.com/paper/2302.12986/) (2023)
- [MovieNet-PS: A Large-Scale Person Search Dataset in the Wild](https://mlsanity.com/paper/2112.02500/) (2021)
- [Deep Intra-Image Contrastive Learning for Weakly Supervised One-Step Person Search](https://mlsanity.com/paper/2302.04607/) (2023)
- [CGUA: Context-Guided and Unpaired-Assisted Weakly Supervised Person Search](https://mlsanity.com/paper/2203.14307/) (2022)
- [Domain Generalizable Person Search Using Unreal Dataset](https://mlsanity.com/paper/2404.00626/) (2024)
- [Prompting Continual Person Search](https://mlsanity.com/paper/2410.19239/) (2024)
- [LEAPS: End-to-End One-Step Person Search With Learnable Proposals](https://mlsanity.com/paper/2303.11859/) (2023)
- [DDAM-PS: Diligent Domain Adaptive Mixer for Person Search](https://mlsanity.com/paper/2310.20706/) (2023)
- [Fast One-Stage Unsupervised Domain Adaptive Person Search](https://mlsanity.com/paper/2405.02832/) (2024)
- [Unsupervised Domain Adaptive Person Search via Dual Self-Calibration](https://mlsanity.com/paper/2412.16506/) (2024)
- [Generalizable Person Search on Open-world User-Generated Video Content](https://mlsanity.com/paper/2310.10068/) (2023)
- [Lifelong Person Search](https://mlsanity.com/paper/2407.21252/) (2024)
- [Swap Path Network for Robust Person Search Pre-training](https://mlsanity.com/paper/2412.05433/) (2024)

## Abstract (arXiv)

Supervised learning is dominant in person search, but it requires elaborate labeling of bounding boxes and identities. Large-scale labeled training data is often difficult to collect, especially for person identities. A natural question is whether a good person search model can be trained without the need of identity supervision. In this paper, we present a weakly supervised setting where only bounding box annotations are available. Based on this new setting, we provide an effective baseline model termed Region Siamese Networks (R-SiamNets). Towards learning useful representations for recognition in the absence of identity labels, we supervise the R-SiamNet with instance-level consistency loss and cluster-level contrastive loss. For instance-level consistency learning, the R-SiamNet is constrained to extract consistent features from each person region with or without out-of-region context. For cluster-level contrastive learning, we enforce the aggregation of closest instances and the separation of dissimilar ones in feature space. Extensive experiments validate the utility of our weakly supervised method. Our model achieves the rank-1 of 87.1% and mAP of 86.0% on CUHK-SYSU benchmark, which surpasses several fully supervised methods, such as OIM and MGTS, by a clear margin. More promising performance can be reached by incorporating extra training data. We hope this work could encourage the future research in this field.