Thinking with Cameras: Active Visual Reasoning via Dynamic Viewpoint Control for Surveillance Video Understanding

• URL: http://arxiv.org/abs/2609.06475v1
• Date: Sun, 06 Sep 2026 08:48:02 GMT
• Title: Thinking with Cameras: Active Visual Reasoning via Dynamic Viewpoint Control for Surveillance Video Understanding
• Abstract summary: We introduce CamVLM, a new framework for thinking with cameras.<n>CamVLM enables LVLMs to actively acquire visual evidence through dynamic viewpoint control.<n>Our datasets, model, and code will be available at https://github.com/xiaozhang79/CamVLM.
• Score: 27.715058333490447
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Large vision-language models (LVLMs) have recently achieved remarkable progress in general-purpose video understanding. However, their application to surveillance videos remains challenging due to the lack of large-scale domain-specific datasets and the limitation of passive observation from fixed viewpoints. In surveillance scenarios, critical visual evidence can be easily missed when targets are distant, small, occluded, or move beyond the current camera view. In this work, we introduce CamVLM, a new framework for Thinking with Cameras, which enables LVLMs to actively acquire visual evidence through dynamic viewpoint control rather than passively analyzing fixed video streams. We first construct CCTV-Anomaly, a large-scale surveillance video understanding dataset containing 14,459 videos across 10 anomaly categories, with detailed captions and event annotations. We further formulate viewpoint control as an active visual perception problem and build CamTrack-53K, an object-centric viewpoint trajectory dataset for learning camera actions. Moreover, we propose a reinforcement learning based viewpoint policy optimization framework, which models camera control as a sequential decision-making process and learns long-horizon observation strategies beyond supervised trajectory imitation. Extensive experiments demonstrate that CamVLM achieves state-of-the-art performance under both passive observation and dynamic viewpoint settings, validating the effectiveness of active camera-based reasoning for surveillance video understanding. Our datasets, model, and code will be available at https://github.com/xiaozhang79/CamVLM .

Related papers

• LIME: Learning Intent-aware Camera Motion from Egocentric Video [56.023644847494836]
  We propose a vision-language camera-motion generator that combines an auto-regressive observation-gain output with a continuous flow-matching pose head.<n>We show that LIME can learn to actively choose camera poses from passive human video, turning ordinary egocentric recordings into supervision for intent-aware active perception.
  arXiv  Detail & Related papers  (2026-07-02T16:48:43Z)
• Warp-as-History: Generalizable Camera-Controlled Video Generation from One Training Video [19.675672131137382]
  We propose a simple interface that turns camera-induced warps into camera-warped pseudo-history.<n>We align its positional encoding with the target frames being denoised and remove warped-history tokens without valid source observations.<n>Our method improves camera adherence, visual quality, and motion dynamics without test-time optimization or target-video adaptation.
  arXiv  Detail & Related papers  (2026-05-14T17:58:26Z)
• Beyond Inpainting: Unleash 3D Understanding for Precise Camera-Controlled Video Generation [21.084121261693365]
  We propose DepthDirector, a video re-rendering framework with precise camera controllability.<n>By leveraging the depth video from explicit 3D representation as camera-control guidance, our method can faithfully reproduce the dynamic scene of an input video under novel camera trajectories.
  arXiv  Detail & Related papers  (2026-01-15T09:26:45Z)
• From Sight to Insight: Unleashing Eye-Tracking in Weakly Supervised Video Salient Object Detection [60.11169426478452]
  This paper aims to introduce fixation information to assist the detection of salient objects under weak supervision.<n>We propose a Position and Semantic Embedding (PSE) module to provide location and semantic guidance during the feature learning process.<n>An Intra-Inter Mixed Contrastive (MCII) model improves thetemporal modeling capabilities under weak supervision.
  arXiv  Detail & Related papers  (2025-06-30T05:01:40Z)
• Object Concepts Emerge from Motion [24.73461163778215]
  We propose a biologically inspired framework for learning object-centric visual representations in an unsupervised manner.<n>Our key insight is that motion boundary serves as a strong signal for object-level grouping.<n>Our framework is fully label-free and does not rely on camera calibration, making it scalable to large-scale unstructured video data.
  arXiv  Detail & Related papers  (2025-05-27T18:09:02Z)
• Towards Understanding Camera Motions in Any Video [89.97247162415158]
  We introduce CameraBench, a large-scale dataset and benchmark designed to assess and improve camera motion understanding.<n>CameraBench consists of 3,000 diverse internet videos annotated by experts through a rigorous quality control process.<n>One of our contributions is a taxonomy of camera motion primitives, designed in collaboration with cinematographers.
  arXiv  Detail & Related papers  (2025-04-21T18:34:57Z)
• C-Drag: Chain-of-Thought Driven Motion Controller for Video Generation [81.4106601222722]
  Trajectory-based motion control has emerged as an intuitive and efficient approach for controllable video generation.<n>We propose a Chain-of-Thought-based motion controller for controllable video generation, named C-Drag.<n>Our method includes an object perception module and a Chain-of-Thought-based motion reasoning module.
  arXiv  Detail & Related papers  (2025-02-27T08:21:03Z)
• Cavia: Camera-controllable Multi-view Video Diffusion with View-Integrated Attention [62.2447324481159]
  Cavia is a novel framework for camera-controllable, multi-view video generation. Our framework extends the spatial and temporal attention modules, improving both viewpoint and temporal consistency. Cavia is the first of its kind that allows the user to specify distinct camera motion while obtaining object motion.
  arXiv  Detail & Related papers  (2024-10-14T17:46:32Z)
• Imitation-Based Active Camera Control with Deep Convolutional Neural Network [4.09920839425892]
  In this paper we frame active visual monitoring as an imitation learning problem to be solved in a supervised manner using deep learning. A deep convolutional neural network is trained end-to-end as the camera controller that learns the entire processing pipeline needed to control a camera to follow multiple targets. Experimental results indicate that the proposed solution is robust to varying conditions and is able to achieve better monitoring performance.
  arXiv  Detail & Related papers  (2020-12-11T15:37:33Z)
• 4D Visualization of Dynamic Events from Unconstrained Multi-View Videos [77.48430951972928]
  We present a data-driven approach for 4D space-time visualization of dynamic events from videos captured by hand-held multiple cameras. Key to our approach is the use of self-supervised neural networks specific to the scene to compose static and dynamic aspects of an event. This model allows us to create virtual cameras that facilitate: (1) freezing the time and exploring views; (2) freezing a view and moving through time; and (3) simultaneously changing both time and view.
  arXiv  Detail & Related papers  (2020-05-27T17:57:19Z)

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.