PulseAugur
实时 16:49:19
English(EN) How Many Visual Levers Drive Urban Perception? Interventional Counterfactuals via Multiple Localised Edits

研究人员探索用于城市感知和安全判断的视觉杠杆

研究人员开发了一个新框架,用于识别影响人类对城市景观感知的特定视觉元素,超越了简单的相关性。这种干预性反事实方法系统地测试了图像的局部编辑(例如,出行基础设施或物理维护的变化)如何改变预测的安全判断。该框架旨在通过生成和验证真实且合理的反事实编辑,来提供对场景可解释性的更深入理解,并以人类判断作为最终验证。 AI

影响 引入了一种新颖的方法来理解特定的视觉变化如何影响AI对城市环境的感知,从而可能提高模型的可解释性。

排序理由 学术论文,介绍了一种用于计算机视觉可解释性的新框架。

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 →

研究人员探索用于城市感知和安全判断的视觉杠杆

报道来源 [2]

  1. arXiv cs.CV TIER_1 English(EN) · Jason Tang, Stephen Law ·

    How Many Visual Levers Drive Urban Perception? Interventional Counterfactuals via Multiple Localised Edits

    arXiv:2604.22103v1 Announce Type: cross Abstract: Street-view perception models predict subjective attributes such as safety at scale, but remain correlational: they do not identify which localized visual changes would plausibly shift human judgement for a specific scene. We prop…

  2. arXiv cs.CV TIER_1 English(EN) · Stephen Law ·

    How Many Visual Levers Drive Urban Perception? Interventional Counterfactuals via Multiple Localised Edits

    Street-view perception models predict subjective attributes such as safety at scale, but remain correlational: they do not identify which localized visual changes would plausibly shift human judgement for a specific scene. We propose a lever-based interventional counterfactual fr…