CrossView: Can Vision-Language Models Reason Across Cameras?

• URL: http://arxiv.org/abs/2608.15539v1
• Date: Sun, 16 Aug 2026 05:25:04 GMT
• Title: CrossView: Can Vision-Language Models Reason Across Cameras?
• Abstract summary: We introduce CrossView, a benchmark for multi-camera video understanding.<n>It spans autonomous driving, security surveillance, egocentric/exocentric video, and robotics.<n> evaluation reveals consistently low accuracy, with open-source models trailing by a wide margin.
• Score: 7.6121119612574155
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Video understanding benchmarks have long centered on single-camera settings, where modern multi-modal language models achieve strong performance across image and video tasks. Yet, the real world runs on multi-camera networks: autonomous vehicles, security systems, and robots all gather data across many simultaneous views. We argue that this is not simply "more" of the single-camera problem; it is fundamentally different. Multi-camera reasoning requires handling context that scales with the number of views, resolving occlusions visible from only a subset of cameras, judging which views matter, and integrating evidence across perspectives that may overlap or diverge. Current models struggle with exactly these challenges, yet no benchmark systematically targets them. We introduce CrossView, a multi-camera video question-answering benchmark spanning autonomous driving, security surveillance, egocentric/exocentric video, and robotics. Evaluation of proprietary models, such as GPT-5.2, and open-source models, like Qwen3-VL, reveals consistently low accuracy, with open-source models trailing by a wide margin. Performance scales strongly with a model's ability to jointly process multiple viewpoints, positioning CrossView as a rigorous benchmark for multi-camera video. We open-source our code and dataset at https://utaustin-swarmlab.github.io/CrossView.

Related papers

• VideoWeaver: Multimodal Multi-View Video-to-Video Transfer for Embodied Agents [17.66237759970927]
  VideoWeaver is the first multimodal multi-view V2V translation framework.<n>We train views at distinct diffusion timesteps, enabling the model to learn both joint and conditional view distributions.<n>Experiments show superior or similar performance to the state-of-the-art on the single-view translation benchmarks.
  arXiv  Detail & Related papers  (2026-03-26T13:14:13Z)
• OmniView: An All-Seeing Diffusion Model for 3D and 4D View Synthesis [80.3346344429389]
  We introduce OmniView, a unified framework that generalizes across a wide range of 4D consistency tasks.<n>Our method separately represents space, time, and view conditions, enabling flexible combinations of these inputs.<n>For example, OmniView can synthesize novel views from static, dynamic, and multiview inputs, extrapolate trajectories forward and backward in time, and create videos from text or image prompts with full camera control.
  arXiv  Detail & Related papers  (2025-12-11T18:59:05Z)
• Multi-View 3D Point Tracking [67.21282192436031]
  We introduce the first data-driven multi-view 3D point tracker, designed to track arbitrary points in dynamic scenes using multiple camera views.<n>Our model directly predicts 3D correspondences using a practical number of cameras.<n>We train on 5K synthetic multi-view Kubric sequences and evaluate on two real-world benchmarks.
  arXiv  Detail & Related papers  (2025-08-28T17:58:20Z)
• Object-centric Video Question Answering with Visual Grounding and Referring [43.963739052764595]
  We introduce a VideoLLM model, capable of performing both object referring for input and grounding for output in video reasoning tasks.<n>We also propose STOM, a novel approach that propagates arbitrary visual prompts input at any single timestamp to the remaining frames within a video.<n>We conduct comprehensive experiments on VideoInfer and other existing benchmarks across video question answering and referring object segmentation.
  arXiv  Detail & Related papers  (2025-07-25T18:11:23Z)
• ReCamMaster: Camera-Controlled Generative Rendering from A Single Video [72.42376733537925]
  ReCamMaster is a camera-controlled generative video re-rendering framework.<n>It reproduces the dynamic scene of an input video at novel camera trajectories.<n>Our method also finds promising applications in video stabilization, super-resolution, and outpainting.
  arXiv  Detail & Related papers  (2025-03-14T17:59:31Z)
• SynCamMaster: Synchronizing Multi-Camera Video Generation from Diverse Viewpoints [43.14498014617223]
  We propose a plug-and-play module that enhances a pre-trained text-to-video model for multi-camera video generation.<n>We introduce a multi-view synchronization module to maintain appearance and geometry consistency across different viewpoints.<n>Our method enables intriguing extensions, such as re-rendering a video from novel viewpoints.
  arXiv  Detail & Related papers  (2024-12-10T18:55:17Z)
• Which Viewpoint Shows it Best? Language for Weakly Supervising View Selection in Multi-view Instructional Videos [66.1935609072708]
  LangView is a framework that uses the relative accuracy of view-dependent caption predictions as a proxy for best view pseudo-labels.<n>During inference, our model takes as input only a multi-view video--no language or camera poses--and returns the best viewpoint to watch at each timestep.
  arXiv  Detail & Related papers  (2024-11-13T16:31:08Z)
• Cross-View Cross-Scene Multi-View Crowd Counting [56.83882084112913]
  Multi-view crowd counting has been previously proposed to utilize multi-cameras to extend the field-of-view of a single camera. We propose a cross-view cross-scene (CVCS) multi-view crowd counting paradigm, where the training and testing occur on different scenes with arbitrary camera layouts.
  arXiv  Detail & Related papers  (2022-05-03T15:03:44Z)
• OmniDet: Surround View Cameras based Multi-task Visual Perception Network for Autonomous Driving [10.3540046389057]
  This work presents a multi-task visual perception network on unrectified fisheye images. It consists of six primary tasks necessary for an autonomous driving system. We demonstrate that the jointly trained model performs better than the respective single task versions.
  arXiv  Detail & Related papers  (2021-02-15T10:46:24Z)
• Wide-Area Crowd Counting: Multi-View Fusion Networks for Counting in Large Scenes [50.744452135300115]
  We propose a deep neural network framework for multi-view crowd counting. Our methods achieve state-of-the-art results compared to other multi-view counting baselines.
  arXiv  Detail & Related papers  (2020-12-02T03:20:30Z)

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.