# Unleashing MLLMs on the Edge: A Unified Framework for Cross-Modal ReID via Adaptive SVD Distillation

arXiv: https://arxiv.org/abs/2602.12936 · HTML: https://mlsanity.com/paper/2602.12936/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 28 of its 51 references that are in the map.
- First posted on arXiv 13 Feb 2026.

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

- [Learning Transferable Visual Models From Natural Language Supervision](https://mlsanity.com/paper/2103.00020/) (2021)
- [LoRA: Low-Rank Adaptation of Large Language Models](https://mlsanity.com/paper/2106.09685/) (2021)
- [Visual Instruction Tuning](https://mlsanity.com/paper/2304.08485/) (2023)
- [Qwen2.5-VL Technical Report](https://mlsanity.com/paper/2502.13923/) (2025)
- [TransReID: Transformer-based Object Re-Identification](https://mlsanity.com/paper/2102.04378/) (2021)
- [Cross-Modal Implicit Relation Reasoning and Aligning for Text-to-Image Person Retrieval](https://mlsanity.com/paper/2303.12501/) (2023)
- [Semantically Self-Aligned Network for Text-to-Image Part-aware Person Re-identification](https://mlsanity.com/paper/2107.12666/) (2021)
- [DSSL: Deep Surroundings-person Separation Learning for Text-based Person Retrieval](https://mlsanity.com/paper/2109.05534/) (2021)
- [Cross-modality Person re-identification with Shared-Specific Feature Transfer](https://mlsanity.com/paper/2002.12489/) (2020)
- [Diverse Embedding Expansion Network and Low-Light Cross-Modality Benchmark for Visible-Infrared Person Re-identification](https://mlsanity.com/paper/2303.14481/) (2023)
- [ViTAA: Visual-Textual Attributes Alignment in Person Search by Natural Language](https://mlsanity.com/paper/2005.07327/) (2020)
- [See Finer, See More: Implicit Modality Alignment for Text-based Person Retrieval](https://mlsanity.com/paper/2208.08608/) (2022)
- [Learning Granularity-Unified Representations for Text-to-Image Person Re-identification](https://mlsanity.com/paper/2207.07802/) (2022)
- [Self-supervised Knowledge Distillation Using Singular Value Decomposition](https://mlsanity.com/paper/1807.06819/) (2018)
- [Learning Semantic-Aligned Feature Representation for Text-based Person Search](https://mlsanity.com/paper/2112.06714/) (2021)
- [CAIBC: Capturing All-round Information Beyond Color for Text-based Person Retrieval](https://mlsanity.com/paper/2209.05773/) (2022)
- [Harnessing the Power of MLLMs for Transferable Text-to-Image Person ReID](https://mlsanity.com/paper/2405.04940/) (2024)
- [Look Before You Leap: Improving Text-based Person Retrieval by Learning A Consistent Cross-modal Common Manifold](https://mlsanity.com/paper/2209.06209/) (2022)
- [EVE: Efficient Vision-Language Pre-training with Masked Prediction and Modality-Aware MoE](https://mlsanity.com/paper/2308.11971/) (2023)
- [Teaching Where to Look: Attention Similarity Knowledge Distillation for Low Resolution Face Recognition](https://mlsanity.com/paper/2209.14498/) (2022)
- [All in One Framework for Multimodal Re-identification in the Wild](https://mlsanity.com/paper/2405.04741/) (2024)
- [When Large Vision-Language Models Meet Person Re-Identification](https://mlsanity.com/paper/2411.18111/) (2024)
- [CLIP-based Synergistic Knowledge Transfer for Text-based Person Retrieval](https://mlsanity.com/paper/2309.09496/) (2023)
- [Open-Qwen2VL: Compute-Efficient Pre-Training of Fully-Open Multimodal LLMs on Academic Resources](https://mlsanity.com/paper/2504.00595/) (2025)
- [MLLMReID: Multimodal Large Language Model-based Person Re-identification](https://mlsanity.com/paper/2401.13201/) (2024)
- [Modeling Thousands of Human Annotators for Generalizable Text-to-Image Person Re-identification](https://mlsanity.com/paper/2503.09962/) (2025)
- [StyleGAN3: Generative Networks for Improving the Equivariance of Translation and Rotation](https://mlsanity.com/paper/2307.03898/) (2023)
- [FlexiReID: Adaptive Mixture of Expert for Multi-Modal Person Re-Identification](https://mlsanity.com/paper/2510.15595/) (2025)

## Abstract (arXiv)

Practical cloud-edge deployment of Cross-Modal Re-identification (CM-ReID) faces challenges due to maintaining a fragmented ecosystem of specialized cloud models for diverse modalities. While Multi-Modal Large Language Models (MLLMs) offer strong unification potential, existing approaches fail to adapt them into a single end-to-end backbone and lack effective knowledge distillation strategies for edge deployment. To address these limitations, we propose MLLMEmbed-ReID, a unified framework based on a powerful cloud-edge architecture. First, we adapt a foundational MLLM into a state-of-the-art cloud model. We leverage instruction-based prompting to guide the MLLM in generating a unified embedding space across RGB, infrared, sketch, and text modalities. This model is then trained efficiently with a hierarchical Low-Rank Adaptation finetuning (LoRA-SFT) strategy, optimized under a holistic cross-modal alignment objective. Second, to deploy its knowledge onto an edge-native student, we introduce a novel distillation strategy motivated by the low-rank property in the teacher's feature space. To prioritize essential information, this method employs a Principal Component Mapping loss, while relational structures are preserved via a Feature Relation loss. Our lightweight edge-based model achieves state-of-the-art performance on multiple visual CM-ReID benchmarks, while its cloud-based counterpart excels across all CM-ReID benchmarks. The MLLMEmbed-ReID framework thus presents a complete and effective solution for deploying unified MLLM-level intelligence on resource-constrained devices. The code and models will be open-sourced soon.