Rethinking Multi-Branch and Cross-Backbone Fusion for Vehicle Re-Identification in the Foundation-Model Era

• URL: http://arxiv.org/abs/2607.22068v1
• Date: Fri, 24 Jul 2026 08:10:53 GMT
• Title: Rethinking Multi-Branch and Cross-Backbone Fusion for Vehicle Re-Identification in the Foundation-Model Era
• Abstract summary: Multi-branch architectures and CNN-Transformer fusion have long been regarded as effective ways to improve vehicle re-identification (Re-ID)<n>A single DINOv3-pretrained ConvNeXt trained with a tuned recipe achieves 88.19 mAP on VeRi-Wild Small and 77.47 mAP on VeRi-Wild Large.<n>We further study cross-backbone fusion using an asymmetric frozen-anchor strategy to combine ConvNeXt and Vision Transformer representations.
• Score: 22.36803510099264
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Multi-branch architectures and CNN-Transformer fusion have long been regarded as effective ways to improve vehicle re-identification (Re-ID) by combining complementary representations. In this work, we revisit this assumption in the foundation-model era through a comprehensive empirical study. A single DINOv3-pretrained ConvNeXt trained with a tuned recipe achieves 88.19 mAP on VeRi-Wild Small and 77.47 mAP on VeRi-Wild Large using visual cues alone, matching the strongest protocol-verified metadata-dependent multi-branch baseline. Applying training-free re-ranking further improves performance to 92.38 and 83.68 mAP, respectively. Using this strong baseline together with retrieval-level branch diagnostics, we evaluate whether increasing representational diversity still provides measurable gains. Across both benchmarks, concatenating multiple branches built on a shared backbone changes the best single-branch performance by less than one mAP point while increasing the embedding dimension by 4x, and the resulting representation has an effective rank close to the original feature dimension. We further study cross-backbone fusion using an asymmetric frozen-anchor strategy to combine ConvNeXt and Vision Transformer representations. Despite these favorable conditions, Transformer branches consistently remain 13-15 mAP below the ConvNeXt backbone, and paired per-query bootstrap analysis estimates the largest observed fusion gain to be only +0.11 mAP (95% confidence interval). Our results suggest that, under the evaluated setting, improving a single strong foundation-model backbone together with retrieval-stage re-ranking is more effective than increasing architectural complexity through additional branches or heterogeneous backbones. We restrict our conclusions to single-seed training and one family of foundation models and discuss conditions under which these observations may not hold.

Related papers

• Mitigating Strong-Modality Collapse in Multimodal Learning via Inverted Asymmetric Fusion [0.0]
  We propose Inverted Asymmetric Fusion, which avoids forcing mutual attention across modalities.<n>We evaluate IAF on three benchmarks with different modality hierarchies: text-dominant datasets (MultiHuSE, UR-FUNNY) and an audio-visual-dominant dataset (MUStARD)
  arXiv  Detail & Related papers  (2026-08-27T09:39:09Z)
• Low Variance Trust Region Optimization with Independent Actors and Sequential Updates in Cooperative Multi-agent Reinforcement Learning [5.8010446129208155]
  We introduce a clipping objective to control the upper bounds of the advantage fluctuation in sequential updates.<n>We derive two new practical algorithms using our clipping objective.<n>Our proposed method is highlighted with both stable convergence properties and the desired low advantage variance estimation.
  arXiv  Detail & Related papers  (2026-06-24T08:05:11Z)
• Multimodal Evaluator Preference Collapse: Cross-Modal Contagion in Self-Evolving Agents [1.0152838128195467]
  We show that Evaluator Preference Collapse ( EPC) is dramatically amplified in multimodal settings.<n>We then demonstrate a novel phenomenon we term cross-modal contagion.<n>We identify cross-model evaluator architecture as the primary risk factor for preference drift.
  arXiv  Detail & Related papers  (2026-06-15T13:18:20Z)
• MM-Doc-R1: Training Agents for Long Document Visual Question Answering through Multi-turn Reinforcement Learning [74.07254720088926]
  We introduce MM-Doc-R1, a novel framework that employs an agentic, vision-aware workflow to address long document visual question answering.<n>We propose Similarity-based Policy Optimization (SPO), addressing baseline estimation bias in existing multi-turn reinforcement learning (RL) algorithms like GRPO.<n>Our experiments on the MMLongbench-Doc benchmark show that MM-Doc-R1 outperforms previous baselines by 10.4%.
  arXiv  Detail & Related papers  (2026-04-15T07:39:08Z)
• BayesRAG: Probabilistic Mutual Evidence Corroboration for Multimodal Retrieval-Augmented Generation [33.53566598271416]
  BayesRAG is a novel multimodal retrieval framework grounded in Bayesian inference and Dempster-Shafer evidence theory.<n>We show that BayesRAG significantly outperforms state-of-the-art (SOTA) methods on challenging multimodal benchmarks.
  arXiv  Detail & Related papers  (2026-01-12T08:53:14Z)
• VAPO: Efficient and Reliable Reinforcement Learning for Advanced Reasoning Tasks [49.0793012627959]
  We present VAPO, a novel framework tailored for reasoning models within the value-based paradigm.<n>VAPO attains a state-of-the-art score of $mathbf60.4$.<n>In direct comparison under identical experimental settings, VAPO outperforms the previously reported results of DeepSeek-R1-Zero-Qwen-32B and DAPO by more than 10 points.
  arXiv  Detail & Related papers  (2025-04-07T14:21:11Z)
• ALoRE: Efficient Visual Adaptation via Aggregating Low Rank Experts [71.91042186338163]
  ALoRE is a novel PETL method that reuses the hypercomplex parameterized space constructed by Kronecker product to Aggregate Low Rank Experts.<n>Thanks to the artful design, ALoRE maintains negligible extra parameters and can be effortlessly merged into the frozen backbone.
  arXiv  Detail & Related papers  (2024-12-11T12:31:30Z)
• SiReRAG: Indexing Similar and Related Information for Multihop Reasoning [96.60045548116584]
  SiReRAG is a novel RAG indexing approach that explicitly considers both similar and related information.<n>SiReRAG consistently outperforms state-of-the-art indexing methods on three multihop datasets.
  arXiv  Detail & Related papers  (2024-12-09T04:56:43Z)
• An Empirical Study of Multimodal Model Merging [148.48412442848795]
  Model merging is a technique that fuses multiple models trained on different tasks to generate a multi-task solution. We conduct our study for a novel goal where we can merge vision, language, and cross-modal transformers of a modality-specific architecture. We propose two metrics that assess the distance between weights to be merged and can serve as an indicator of the merging outcomes.
  arXiv  Detail & Related papers  (2023-04-28T15:43:21Z)
• Learnable Pillar-based Re-ranking for Image-Text Retrieval [119.9979224297237]
  Image-text retrieval aims to bridge the modality gap and retrieve cross-modal content based on semantic similarities. Re-ranking, a popular post-processing practice, has revealed the superiority of capturing neighbor relations in single-modality retrieval tasks. We propose a novel learnable pillar-based re-ranking paradigm for image-text retrieval.
  arXiv  Detail & Related papers  (2023-04-25T04:33:27Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.