基于 RGB-事件的行人属性识别：一个基准数据集与一个非对称 RWKV 融合框架
RGB-Event based Pedestrian Attribute Recognition: A Benchmark Dataset and An Asymmetric RWKV Fusion Framework
提供了首个大规模 RGB-Event 配对的行人属性识别数据集与基准，并给出基于 RWKV 的非对称融合基线，可直接用于研究事件相机与 RGB 的跨模态融合、低光照与运动模糊场景下的视觉识别。
原文依据 we introduce the first large-scale multi-modal pedestrian attribute recognition dataset, termed EventPAR, comprising 100K paired RGB-Event samples
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论现有的行人属性识别方法通常基于 RGB 帧相机开发。然而，这些方法受限于 RGB 相机的固有缺陷，例如对光照条件敏感、易受运动模糊影响，从而制约了其性能。此外，当前属性识别主要聚焦于分析行人的外部外观与衣着，缺乏对情感维度的探索。本文重新审视这些问题，借鉴事件相机在低光照、高速与低功耗方面的优势，提出一种新的多模态 RGB-事件属性识别任务。具体而言，我们引入了首个大规模多模态行人属性识别数据集，命名为 EventPAR，包含 10 万对 RGB-事件样本，覆盖与外观及六种人类情感相关的 50 种属性、多样场景与多种季节。通过在该数据集上重新训练并评估主流 PAR 模型，我们建立了全面的基准，并在数据与算法基线层面为未来研究提供了坚实基础。此外，我们提出一种新的基于 RWKV 的多模态行人属性识别框架，包含一个 RWKV 视觉编码器与一个非对称 RWKV 融合模块。在我们提出的数据集以及两个模拟数据集（MARS-Attribute 与 DukeMTMC-VID-Attribute）上开展了大量实验，取得了最优结果。源代码与数据集将在论文录用后发布。
英文原文摘要 1543 字符
Existing pedestrian attribute recognition methods are generally developed based on RGB frame cameras. However, these approaches are constrained by the limitations of RGB cameras, such as sensitivity to lighting conditions and motion blur, which hinder their performance. Furthermore, current attribute recognition primarily focuses on analyzing pedestrians' external appearance and clothing, lacking an exploration of emotional dimensions. In this paper, we revisit these issues and propose a novel multi-modal RGB-Event attribute recognition task by drawing inspiration from the advantages of event cameras in low-light, high-speed, and low-power consumption. Specifically, we introduce the first large-scale multi-modal pedestrian attribute recognition dataset, termed EventPAR, comprising 100K paired RGB-Event samples that cover 50 attributes related to both appearance and six human emotions, diverse scenes, and various seasons. By retraining and evaluating mainstream PAR models on this dataset, we establish a comprehensive benchmark and provide a solid foundation for future research in terms of data and algorithmic baselines. In addition, we propose a novel RWKV-based multi-modal pedestrian attribute recognition framework, featuring an RWKV visual encoder and an asymmetric RWKV fusion module. Extensive experiments are conducted on our proposed dataset as well as two simulated datasets (MARS-Attribute and DukeMTMC-VID-Attribute), achieving state-of-the-art results. The source code and dataset will be released upon acceptance.