Lune

ACL Findings2026

Bridging the Pose-Semantic Gap: A Cascade Framework for Text-Based Person Anomaly Search

Zequn Xie, Guijin Luo, Chuxin Wang, Sihang Cai, Tao Jin, Zhou Zhao, Yixuan Tang

Abstract

Text-based person anomaly search retrieves specific behavioral events from surveillance archives using natural-language queries. Although recent pose-aware methods align geometric structures well, they face a fundamental Pose-Semantic Gap: semantically different actions can share similar skeletal geometries. While Multimodal Large Language Models (MLLMs) can reduce this ambiguity, using them for large-scale retrieval is computationally prohibitive. We propose the Structure-Semantic Decoupled Cascade (SSDC) framework, which decouples retrieval into two stages:

(1) Structure-Aware Coarse Retrieval, where a lightweight model quickly filters candidates by skeletal similarity; and (2) Detective Squad Interaction, a multi-agent semantic verification module. The squad consists of a Detective for fast binary filtering, an Analyst for evidence extraction, and a Writer for semantic synthesis. Finally, we re-rank candidates by fusing the synthesized captions with structural priors. Experiments on the PAB benchmark show that SSDC achieves state-of-theart performance by balancing efficiency and semantic reasoning.Our code is available at: https://github.com/GridNexus/SSDC.

Ask Lune

Know this paper in minutes, not an afternoon.

Lune can read every page of it and the top-tier papers that cite it, and gives you its claims, its evidence and its limits, each one cited to the paper it comes from. No skimming, no AI slop.

Questions to start from

Lune can draw on

the full paper · related work

Get the answer

Free to start. No credit card required.

Lineage

  • DSSL: Deep Surroundings-person Separation Learning for Text-based Person Retrieval cites Pose-Guided Multi-Granularity Attention Network for Text-Based Person Search
  • Multi-Grained Vision Language Pre-Training: Aligning Texts with Visual Concepts cites Learning Transferable Visual Models From Natural Language Supervision
Builds onCited byLeft to right by year; the more cited, the higher.

Builds on13

  1. Learning Transferable Visual Models From Natural Language SupervisionAlec Radford, Jong Wook Kim, Chris Hallacy, Aditya Ramesh et al.ICML 2021 · 47,906 citations
  2. Multi-Grained Vision Language Pre-Training: Aligning Texts with Visual ConceptsYan Zeng, Xinsong Zhang, Hang LiICML 2022 · 371 citations
  3. DSSL: Deep Surroundings-person Separation Learning for Text-based Person RetrievalAichun Zhu, Zijie Wang, Yifeng Li, Xili Wan et al.ACM MM 2021 · 274 citations
  4. Generative Cooperative Learning for Unsupervised Video Anomaly DetectionMuhammad Zaigham Zaheer, Arif Mahmood, Muhammad Haris Khan, Mattia Segù et al.CVPR 2022 · 195 citations
  5. Pose-Guided Multi-Granularity Attention Network for Text-Based Person SearchYa Jing, Chenyang Si, Junbo Wang, Wei Wang et al.AAAI 2020 · 182 citations

Related papers

  1. Beyond Walking: A Large-Scale Image-Text Benchmark for Text-Based Person Anomaly SearchShuyu Yang, Yaxiong Wang, Li Zhu, Zhedong ZhengICCV 2025 · 1 citation
  2. Can Multimodal LLMs Perform Time Series Anomaly Detection?Xiongxiao Xu, Haoran Wang, Yueqing Liang, Philip S. Yu et al.WWW 2026 · 18 citations
  3. Decompose and Conquer: Compositional Reasoning for Zero-Shot Temporal Action LocalizationHaoyu Tang, Tianyuan Liang, Han Jiang, Xuesong Liu et al.AAAI 2026
  4. Chat-Driven Text Generation and Interaction for Person RetrievalZequn Xie, Chuxin Wang, Yeqiang Wang, Sihang Cai et al.EMNLP 2025 · 12 citations
  5. Beyond Label Semantics:Language-Guided Action Anatomy for Few-Shot Action RecognitionZefeng Qian, Xincheng Yao, Yifei Huang, Chongyang Zhang et al.ICCV 2025 · 4 citations
Dusk over the sea between two cliffs drawn in fine vertical lines