paper-with-me

홈 › Papers

Cross-Camera View-Overlap Recognition

2022-08-24 · Alessio Xompero, Andrea Cavallaro

We propose a decentralised view-overlap recognition framework that operates across freely moving cameras without the need of a reference 3D map. Each camera independently extracts, aggregates into a hierarchical structure, and shares feature-point descriptors over time. A view overlap is recognised by view-matching and geometric validation to discard wrongly matched views. The proposed framework is generic and can be used with different descriptors. We conduct the experiments on publicly available sequences as well as new sequences we collected with hand-held cameras. We show that Oriented FAST and Rotated BRIEF (ORB) features with Bags of Binary Words within the proposed framework lead to higher precision and a higher or similar accuracy compared to NetVLAD, RootSIFT, and SuperGlue.

📄 PDF Abstract BibTeX arXiv:2208.11661

Code (0)

등록된 구현이 없습니다.

Similar Papers 제목 키워드 기반

SurroundNEXO: Ego-Centric Metric Bridging for Spatially Consistent Geometry in Autonomous Driving

2026-06-15 · Shuai Yuan, Runxi Tang, Yuzhou Ji, Fudong Ge 외 arxiv

Modern autonomous driving depends on accurate metric 3D understanding for perception, reconstruction, and planning, which in turn requires reliable multi-camera depth prediction. However, the outward-facing nature of veh…

Zero-shot GeneralizationAutonomous Driving

SDGE: Stereo Guided Depth Estimation for 360$^\circ$ Camera Sets

2024-02-19 · Jialei Xu, Wei Yin, Dong Gong, Junjun Jiang 외

Depth estimation is a critical technology in autonomous driving, and multi-camera systems are often used to achieve a 360$^\circ$ perception. These 360$^\circ$ camera sets often have limited or low-quality overlap region…

3D Object DetectionAutonomous DrivingDepth Estimationobject-detection+1

Enabling Cross-Camera Collaboration for Video Analytics on Distributed Smart Cameras

2024-01-25 · Chulhong Min, Juheon Yi, Utku Gunay Acer, Fahim Kawsar

Overlapping cameras offer exciting opportunities to view a scene from different angles, allowing for more advanced, comprehensive and robust analysis. However, existing visual analytics systems for multi-camera streams a…

Object

4D Human-Scene Reconstruction from Low-Overlap Captures

2026-07-10 · Minhyuk Hwang, Sangmin Kim, Seunguk Do, Daneul Kim 외 arxiv

Existing volumetric capture of dynamic human performance achieves high fidelity with dense camera arrays. However, in real-world scenarios, only a handful of low-overlap cameras are available, which degrades the output q…

Novel View Synthesis

Self-Supervised Partial Cycle-Consistency for Multi-View Matching

2025-01-10 · Fedor Taggenbrock, Gertjan Burghouts, Ronald Poppe

Matching objects across partially overlapping camera views is crucial in multi-camera systems and requires a view-invariant feature extraction network. Training such a network with cycle-consistency circumvents the need …

Scene Understanding