Metraipolis › archive › dc_f795d30dff98

SnapBench: Benchmarking Snap-and-Ask Multimodal Retrieval for Mobile Interactions — arXiv:2608.29607

Source: arXiv (en)
Original: https://arxiv.org/abs/2608.29607
License (as archived): CC0-1.0
Archived: 2026-10-17T09:44:21.573Z
SHA-256 of the archived text: c0f0a598d154e905e5f2b2636a843e3ed81beb56bac83f7447363373ed947a2f
Stable URL of this page: https://metraipolis.com/documento/dc_f795d30dff98

How to cite

SnapBench: Benchmarking Snap-and-Ask Multimodal Retrieval for Mobile Interactions — arXiv:2608.29607 — arXiv (en), https://arxiv.org/abs/2608.29607 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_f795d30dff98 (Metraipolis archive dc_f795d30dff98, sha256 c0f0a598d154e905e5f2b2636a843e3ed81beb56bac83f7447363373ed947a2f)

BibTeX

@misc{dcf795d30dff98,
  title = {SnapBench: Benchmarking Snap-and-Ask Multimodal Retrieval for Mobile Interactions — arXiv:2608.29607},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2608.29607},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_f795d30dff98; sha256 c0f0a598d154e905e5f2b2636a843e3ed81beb56bac83f7447363373ed947a2f},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.29607. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Mobile AI acts as a visual oracle, empowering users to snap a picture of something and ask for information. Snap-and-ask retrieval is now one of the most common entry points for mobile AI, yet photos are often blurry, while text questions may be short or mistyped. Existing benchmarks only test on clean inputs or do not isolate paired robustness in snap-and-ask retrieval. Therefore, we introduce SnapBench, the first paired benchmark for robust snap-and-ask multimodal retrieval, spanning 1,145 queries, 9,085 gallery items under 53 controlled corruption conditions with human annotations. We evaluate 16 multimodal retrievers, covering dual-tower encoders and embedding-based VLMs. Results show that image corruptions substantially degrade retrieval, while text corruptions mainly affect text-only retrieval and have limited impact on joint retrieval. Clean image-only retrieval often outperforms joint retrieval, indicating the coarse-text drag and the lack of cross-modal fallback under noisy inputs. SnapBench provides a controlled testbed for evaluating robust retrieval in snap-and-ask scenarios. We further propose MOOR (Modality-anchored, Outlier-aware, Optimal Reweighting), a simple adaptive fusion approach, highlighting the need for reliability-aware modality calibration in snap-and-ask retrieval.

Tema degli esploratori: statistica e scienza dei dati (retro-catalogo arXiv: annunci del 2026-09-01 in cs.IR).
Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV) ; Information Retrieval (cs.IR).
Pagina del lavoro: https://arxiv.org/abs/2608.29607 — 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 2026-10-17T09:12:53.496Z dalla pagina https://arxiv.org/catchup/cs.IR/2026-09-01?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).