# HBReID: Harder Batch for Re-identification

arXiv: https://arxiv.org/abs/2112.04761 · HTML: https://mlsanity.com/paper/2112.04761/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 28 of its 35 references that are in the map.
- First posted on arXiv 9 Dec 2021.

## Builds on (most cited first, 28 of 28)

- [Domain-Adversarial Training of Neural Networks](https://mlsanity.com/paper/1505.07818/) (2015)
- [Adversarial Discriminative Domain Adaptation](https://mlsanity.com/paper/1702.05464/) (2017)
- [Random Erasing Data Augmentation](https://mlsanity.com/paper/1708.04896/) (2017)
- [In Defense of the Triplet Loss for Person Re-Identification](https://mlsanity.com/paper/1703.07737/) (2017)
- [Person Transfer GAN to Bridge Domain Gap for Person Re-Identification](https://mlsanity.com/paper/1711.08565/) (2017)
- [TransReID: Transformer-based Object Re-Identification](https://mlsanity.com/paper/2102.04378/) (2021)
- [Contrastive Adaptation Network for Unsupervised Domain Adaptation](https://mlsanity.com/paper/1901.00976/) (2019)
- [Circle Loss: A Unified Perspective of Pair Similarity Optimization](https://mlsanity.com/paper/2002.10857/) (2020)
- [Multi-Similarity Loss with General Pair Weighting for Deep Metric Learning](https://mlsanity.com/paper/1904.06627/) (2019)
- [Unsupervised Pixel-Level Domain Adaptation with Generative Adversarial Networks](https://mlsanity.com/paper/1612.05424/) (2016)
- [Sampling Matters in Deep Embedding Learning](https://mlsanity.com/paper/1706.07567/) (2017)
- [Re-ranking Person Re-identification with k-reciprocal Encoding](https://mlsanity.com/paper/1701.08398/) (2017)
- [Simultaneous Deep Transfer Across Domains and Tasks](https://mlsanity.com/paper/1510.02192/) (2015)
- [Proxy Anchor Loss for Deep Metric Learning](https://mlsanity.com/paper/2003.13911/) (2020)
- [Correlation Alignment for Unsupervised Domain Adaptation](https://mlsanity.com/paper/1612.01939/) (2016)
- [Partial Adversarial Domain Adaptation](https://mlsanity.com/paper/1808.04205/) (2018)
- [Why Can't I Dance in the Mall? Learning to Mitigate Scene Bias in Action Recognition](https://mlsanity.com/paper/1912.05534/) (2019)
- [An introduction to domain adaptation and transfer learning](https://mlsanity.com/paper/1812.11806/) (2018)
- [ProxyNCA++: Revisiting and Revitalizing Proxy Neighborhood Component Analysis](https://mlsanity.com/paper/2004.01113/) (2020)
- [Ranked List Loss for Deep Metric Learning](https://mlsanity.com/paper/1903.03238/) (2019)
- [Learning Deep Embeddings with Histogram Loss](https://mlsanity.com/paper/1611.00822/) (2016)
- [The Devil is in the Details: Self-Supervised Attention for Vehicle Re-Identification](https://mlsanity.com/paper/2004.06271/) (2020)
- [Fine-grained Recognition in the Wild: A Multi-Task Domain Adaptation Approach](https://mlsanity.com/paper/1709.02476/) (2017)
- [Towards Discriminative Representation Learning for Unsupervised Person Re-identification](https://mlsanity.com/paper/2108.03439/) (2021)
- [FlipReID: Closing the Gap between Training and Inference in Person Re-Identification](https://mlsanity.com/paper/2105.05639/) (2021)
- [A Person Re-identification Data Augmentation Method with Adversarial Defense Effect](https://mlsanity.com/paper/2101.08783/) (2021)
- [OH-Former: Omni-Relational High-Order Transformer for Person Re-Identification](https://mlsanity.com/paper/2109.11159/) (2021)
- [Graph Convolution for Re-ranking in Person Re-identification](https://mlsanity.com/paper/2107.02220/) (2021)

## Abstract (arXiv)

Triplet loss is a widely adopted loss function in ReID task which pulls the hardest positive pairs close and pushes the hardest negative pairs far away. However, the selected samples are not the hardest globally, but the hardest only in a mini-batch, which will affect the performance. In this report, a hard batch mining method is proposed to mine the hardest samples globally to make triplet harder. More specifically, the most similar classes are selected into a same mini-batch so that the similar classes could be pushed further away. Besides, an adversarial scene removal module composed of a scene classifier and an adversarial loss is used to learn scene invariant feature representations. Experiments are conducted on dataset MSMT17 to prove the effectiveness, and our method surpasses all of the previous methods and sets state-of-the-art result.