![[2508.14746] MissionHD: Hyperdimensional Refinement of Distribution-Deficient Reasoning Graphs for Video Anomaly Detection](https://arxiv.org/static/browse/0.3.4/images/arxiv-logo-fb.png){kind=logo}
[2508.14746] MissionHD: Hyperdimensional Refinement of Distribution-Deficient Reasoning Graphs for Video Anomaly Detection
Summary
The paper presents MissionHD, a novel approach for video anomaly detection using hyperdimensional refinement of reasoning graphs, addressing limitations of traditional methods.
Why It Matters
As video anomaly detection becomes critical in various applications, this research introduces a new paradigm that enhances the effectiveness of reasoning graphs, potentially improving detection accuracy in real-world scenarios. The approach circumvents the limitations of conventional methods by optimizing graph structures directly in hyperdimensional space.
Key Takeaways
- MissionHD leverages hyperdimensional computing for video anomaly detection.
- The method optimizes graph representations without relying on distribution modeling.
- Performance improvements are demonstrated on benchmark datasets, indicating practical applicability.
Computer Science > Machine Learning arXiv:2508.14746 (cs) [Submitted on 20 Aug 2025 (v1), last revised 13 Feb 2026 (this version, v4)] Title:MissionHD: Hyperdimensional Refinement of Distribution-Deficient Reasoning Graphs for Video Anomaly Detection Authors:Sanggeon Yun, Raheeb Hassan, Ryozo Masukawa, Nathaniel D. Bastian, Mohsen Imani View a PDF of the paper titled MissionHD: Hyperdimensional Refinement of Distribution-Deficient Reasoning Graphs for Video Anomaly Detection, by Sanggeon Yun and 4 other authors View PDF HTML (experimental) Abstract:LLM-generated reasoning graphs, referred to as mission-specific graphs (MSGs), are increasingly used for video anomaly detection (VAD) and recognition (VAR). However, they are typically treated as fixed despite being generic and distribution-deficient. Conventional graph structure refinement (GSR) methods are ill-suited to this setting, as they rely on learning structural distributions that are absent in LLM-generated graphs. We propose HDC-constrained Graph Structure Refinement (HDC-GSR), a new paradigm that directly optimizes a decodable, task-aligned graph representation in a single hyperdimensional space without distribution modeling. Leveraging Hyperdimensional Computing (HDC), our framework encodes graphs via binding and bundling operations, aligns the resulting graph code with downstream loss, and decodes edge contributions to refine the structure. We instantiate this approach as MissionHD for weakly supervised VAD/VAR an...
