# Beyond Discrete Samples: High Information Density Replay for Efficient Lifelong Person Re-Identification

arXiv: https://arxiv.org/abs/2508.01587 · HTML: https://mlsanity.com/paper/2508.01587/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 22 of its 52 references that are in the map.
- First posted on arXiv 3 Aug 2025.

## Builds on (most cited first, 22 of 22)

- [Arbitrary Style Transfer in Real-time with Adaptive Instance Normalization](https://mlsanity.com/paper/1703.06868/) (2017)
- [Performance Measures and a Data Set for Multi-Target, Multi-Camera Tracking](https://mlsanity.com/paper/1609.01775/) (2016)
- [Domain Generalization with MixStyle](https://mlsanity.com/paper/2104.02008/) (2021)
- [Deep Learning for Person Re-identification: A Survey and Outlook](https://mlsanity.com/paper/2001.04193/) (2020)
- [Person Transfer GAN to Bridge Domain Gap for Person Re-Identification](https://mlsanity.com/paper/1711.08565/) (2017)
- [Dataset Condensation with Distribution Matching](https://mlsanity.com/paper/2110.04181/) (2021)
- [A Strong Baseline and Batch Normalization Neck for Deep Person Re-identification](https://mlsanity.com/paper/1906.08332/) (2019)
- [Generalizing Dataset Distillation via Deep Generative Prior](https://mlsanity.com/paper/2305.01649/) (2023)
- [Lifelong Person Re-Identification via Adaptive Knowledge Accumulation](https://mlsanity.com/paper/2103.12462/) (2021)
- [Lifelong Person Re-Identification via Knowledge Refreshing and Consolidation](https://mlsanity.com/paper/2211.16201/) (2022)
- [Continual Learning of Diffusion Models with Generative Distillation](https://mlsanity.com/paper/2311.14028/) (2023)
- [Summarizing Stream Data for Memory-Constrained Online Continual Learning](https://mlsanity.com/paper/2305.16645/) (2023)
- [DASK: Distribution Rehearsing via Adaptive Style Kernel Learning for Exemplar-Free Lifelong Person Re-Identification](https://mlsanity.com/paper/2412.09224/) (2024)
- [Video-based Visible-Infrared Person Re-Identification with Auxiliary Samples](https://mlsanity.com/paper/2311.15571/) (2023)
- [Domain Generalization with Correlated Style Uncertainty](https://mlsanity.com/paper/2212.09950/) (2022)
- [Group Distributionally Robust Dataset Distillation with Risk Minimization](https://mlsanity.com/paper/2402.04676/) (2024)
- [Color Prompting for Data-Free Continual Unsupervised Domain Adaptive Person Re-Identification](https://mlsanity.com/paper/2308.10716/) (2023)
- [An Open-World, Diverse, Cross-Spatial-Temporal Benchmark for Dynamic Wild Person Re-Identification](https://mlsanity.com/paper/2403.15119/) (2024)
- [Robust Privacy-Preserving Motion Detection and Object Tracking in Encrypted Streaming Video](https://mlsanity.com/paper/2108.13141/) (2021)
- [A Novel Incremental Learning Driven Instance Segmentation Framework to Recognize Highly Cluttered Instances of the Contraband Items](https://mlsanity.com/paper/2201.02560/) (2022)
- [Progressive Latent Replay for efficient Generative Rehearsal](https://mlsanity.com/paper/2207.01562/) (2022)
- [Synthetic-To-Real Video Person Re-ID](https://mlsanity.com/paper/2402.02108/) (2024)

## Abstract (arXiv)

Lifelong Person Re-Identification (LReID) typically resists catastrophic forgetting by replaying historical samples, rehearsing domain distributions, or distilling previous model knowledge. Among these, data replay is favored for its simplicity and efficiency, as it fundamentally relies on storing discrete raw images. Although often claimed to be efficient, repeatedly training on an accumulating replay buffer with complex selection strategies across sequential domains is actually highly inefficient. Furthermore, this discrete selection severely restricts historical data coverage and results in low information density, inevitably leading to poor generalization on evolving domains and causing these methods to gradually fall behind other approaches. In this paper, we rethink LReID replay and shift the paradigm from sample selection to information compression, proposing a High Information Density Replay (HiDeR) framework. Rather than saving sparse instances, we continually consolidate historical data into a compact, fixed-budget memory. Specifically, we introduce a complexity aware allocation mechanism to dynamically assign memory quotas based on intra-class variance, alongside a metric guided condensation objective that directly preserves essential identity topologies. Furthermore, since highly compressed synthetic samples exhibit artifact styles unsuitable for current domain training, we introduce a cross modality adaptation strategy. By bidirectionally translating styles between synthetic and real samples, this strategy bridges the modality gap to mitigate optimization conflicts, while also enriching stylistic diversity for better generalization. Extensive experiments demonstrate that our framework outperforms state-of-the-art methods in retaining historical knowledge and improving overall generalization, while substantially reducing the cumulative replay cost.