跳到正文
arXiv cs.AI· Olaf Hellwich, Moritz Heine, Patrik Reiske, Sole Traverso, Katharina Hohlbaum, Henning Sprekeler, Christa Th\"one-Reineke, Lars Lewejohann, Roman Bhuiyan·· 4 小时前AI 评分16

基于摄影测量 3D 模型的动物行为分类

Photogrammetric 3D Model-Based Animal Behavior Classification

AI 导读

研究者提出一种两阶段动物行为分类方法:先从视频重建 3D 体部与物体几何,再进行行为分类,而非直接从图像序列端到端估计。该方法让 2D 到 3D 病态逆问题的歧义显式化并可能被纠正。团队先用模拟的交互小鼠验证该策略,并扩展到真实视频。

正文 · 原文

View PDF HTML (experimental)

Abstract:Video-based behavior estimation is often implemented end-to-end from image sequences. We advocate a two-stage alternative: reconstruct 3D body and object geometry before classifying behavior. This makes ambiguities of the ill-posed 2D-to-3D inverse problem explicit and potentially correctable. We demonstrate the strategy with simulated interacting mice and extend it to real video.
Comments: 8 pages, 9 figures, 1 table, conference proceedings
Subjects: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
Cite as: arXiv:2611.17881 [cs.CV]
  (or arXiv:2611.17881v1 [cs.CV] for this version)
  https://doi.org/10.48550/arXiv.2611.17881

arXiv-issued DOI via DataCite (pending registration)

Journal reference: GFaI Future Tech Days Tagungsband 2026, pp. 56 - 63, Gesellschaft zur Förderung angewandter Informatik e.v., Berlin 2026, ISBN 978-3-942709-38-5

Submission history

From: Olaf Hellwich [view email]
[v1] Fri, 13 Nov 2026 08:15:39 UTC (14,021 KB)

来源:arXiv cs.AI · arxiv.org