{"ID":2852895,"CreatedAt":"2026-06-01T04:54:23.091178241Z","UpdatedAt":"2026-06-01T04:54:23.091178241Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2510.17739","arxiv_id":"2510.17739","title":"Joint Multi-Condition Representation Modelling via Matrix Factorisation for Visual Place Recognition","abstract":"We address multi-reference visual place recognition (VPR), where reference sets captured under varying conditions are used to improve localisation performance. While deep learning with large-scale training improves robustness, increasing data diversity and model complexity incur extensive computational cost during training and deployment. Descriptor-level fusion via voting or aggregation avoids training, but often targets multi-sensor setups or relies on heuristics with limited gains under appearance and viewpoint change. We propose a training-free, descriptor-agnostic approach that jointly models places using multiple reference descriptors via matrix decomposition into basis representations, enabling projection-based residual matching. We also introduce SotonMV, a structured benchmark for multi-viewpoint VPR. On multi-appearance data, our method improves Recall@1 by up to ~18% over single-reference and outperforms multi-reference baselines across appearance and viewpoint changes, with gains of ~5% on unstructured data, demonstrating strong generalisation while remaining lightweight.","short_abstract":"We address multi-reference visual place recognition (VPR), where reference sets captured under varying conditions are used to improve localisation performance. While deep learning with large-scale training improves robustness, increasing data diversity and model complexity incur extensive computational cost during trai...","url_abs":"https://arxiv.org/abs/2510.17739","url_pdf":"https://arxiv.org/pdf/2510.17739v1","authors":"[\"Timur Ismagilov\",\"Shakaiba Majeed\",\"Michael Milford\",\"Tan Viet Tuyen Nguyen\",\"Sarvapali D. Ramchurn\",\"Shoaib Ehsan\"]","published":"2025-10-20T16:50:03Z","proceeding":"cs.CV","tasks":"[\"cs.CV\"]","methods":"[]","has_code":false}