Metraipolis › archive › dc_e04e528e07a1

Front-to-Back: Benchmarking Vision-Language Models for Asymmetric Cross-View Vehicle Re-Identification — arXiv:2609.39492

Source
arXiv (en)
Original
https://arxiv.org/abs/2609.39492
License (as archived)
CC0-1.0
Archived
2026-10-07T14:46:09.979Z
SHA-256 of the archived text
f8a8725dc4e928b563ac0ac23204dcfd731eed0c0c50467354fa5af14b2cb6c1
Stable URL of this page
https://metraipolis.com/documento/dc_e04e528e07a1

How to cite

Front-to-Back: Benchmarking Vision-Language Models for Asymmetric Cross-View Vehicle Re-Identification — arXiv:2609.39492 — arXiv (en), https://arxiv.org/abs/2609.39492 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_e04e528e07a1 (Metraipolis archive dc_e04e528e07a1, sha256 f8a8725dc4e928b563ac0ac23204dcfd731eed0c0c50467354fa5af14b2cb6c1)
BibTeX
@misc{dce04e528e07a1,
  title = {Front-to-Back: Benchmarking Vision-Language Models for Asymmetric Cross-View Vehicle Re-Identification — arXiv:2609.39492},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.39492},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_e04e528e07a1; sha256 f8a8725dc4e928b563ac0ac23204dcfd731eed0c0c50467354fa5af14b2cb6c1},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.39492. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Matching the same vehicle across front and rear cameras is difficult because the cameras do not share a view and the vehicle's appearance changes substantially. We introduce Front2Back-ReID, a benchmark of 500 manually verified vehicle handovers from 20 recording sequences in South Africa. Each example asks a model to match a vehicle highlighted in a front-camera image to the same vehicle among at least three candidates in a later rear-camera image. We evaluate seven zero-shot vision-language models, four image-retrieval baselines, and 25 human participants. Models are tested using full front RGB images, cropped target vehicles, and binary silhouettes. The strongest VLM achieved 76.6 percent Rank-1 accuracy on target crops, compared with 74.0 percent for the frozen SigLIP2 baseline; this difference was not statistically clear. Human participants achieved 94.0 percent accuracy with full images and 92.2 percent with target crops. Under our evaluation setup, enabling reasoning improved accuracy across all three input conditions for every model evaluated in both modes. We also found that VLMs generally performed worse on full scenes than on target crops. These results show that general-purpose VLMs do not yet consistently outperform strong visual retrieval for front-to-rear vehicle matching, while humans remain substantially more reliable.

Tema degli esploratori: statistica e scienza dei dati (retro-catalogo arXiv: annunci del 2026-10-01 in cs.CV).
Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV).
Pagina del lavoro: https://arxiv.org/abs/2609.39492 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-10-07T14:19:00.750Z dalla pagina https://arxiv.org/catchup/cs.CV/2026-10-01?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).