PulseAugur
实时 09:15:55
English(EN) Uncertainty-Aware Pedestrian Attribute Recognition via Evidential Deep Learning

新框架使用证据深度学习进行不确定性感知的行人属性识别

研究人员开发了UAPAR,一种用于行人属性识别的新型框架,该框架结合了证据深度学习(EDL)来评估预测的可靠性。与传统的确定性方法不同,这种方法旨在通过识别不可靠的预测来提高复杂环境下的系统鲁棒性。UAPAR利用基于CLIP的架构,并结合了区域感知证据推理模块和证据头来估计属性级别的认知不确定性,并采用不确定性引导的课程学习策略来减轻标签噪声。 AI

影响 通过量化预测不确定性来增强计算机视觉系统的鲁棒性,可能改进安全关键型应用。

排序理由 介绍行人属性识别新框架的学术论文。

在 arXiv cs.CV 阅读 →

AI 生成摘要 · Google Gemini · 来自 2 个来源。 我们如何撰写摘要 →

新框架使用证据深度学习进行不确定性感知的行人属性识别

报道来源 [2]

  1. arXiv cs.CV TIER_1 English(EN) · Zhuofan Lou, Shihang Zhang, Fangle Zhu, Shengjie Ye, Pingyu Wang ·

    Uncertainty-Aware Pedestrian Attribute Recognition via Evidential Deep Learning

    arXiv:2604.26873v1 Announce Type: new Abstract: We propose UAPAR, an Uncertainty-Aware Pedestrian Attribute Recognition framework. To the best of our knowledge, this is the first EDL-based uncertainty-aware framework for pedestrian attribute recognition (PAR). Unlike conventional…

  2. arXiv cs.CV TIER_1 English(EN) · Pingyu Wang ·

    Uncertainty-Aware Pedestrian Attribute Recognition via Evidential Deep Learning

    We propose UAPAR, an Uncertainty-Aware Pedestrian Attribute Recognition framework. To the best of our knowledge, this is the first EDL-based uncertainty-aware framework for pedestrian attribute recognition (PAR). Unlike conventional deterministic methods, which fail to assess pre…