超越扫描路径：动态场景中的图基 gaze 模拟
Beyond Scanpaths: Graph-Based Gaze Simulation in Dynamic Scenes
AI总结：
本文提出基于图的 gaze 模拟方法，通过构建 gaze-中心图和对象密度网络，实现对动态场景中 gaze 动态的精确建模，提供更自然的 gaze 轨迹和 saliency 映射。
摘要：
准确建模人类注意力对许多计算机视觉应用至关重要，尤其是在自动驾驶安全领域。现有方法通常将 gaze 塌缩为 saliency 映射或扫描路径，仅隐式处理 gaze 动态。我们将其建模为自回归动态系统，显式展开时间上的原始 gaze 轨迹，条件于 gaze 历史和变化的环境。驾驶场景被表示为 gaze-中心图，由 Affinity Relation Transformer (ART) 处理，该异构图变压器建模驾驶员 gaze、交通物体和道路结构之间的相互作用。我们进一步引入 Object Density Network (ODN) 来预测下一步 gaze 分布，捕捉复杂环境中注意力转移的随机性和物体中心性质。我们还发布了 Focus100，一个包含 30 名参与者观看第一人称驾驶视频的原始 gaze 数据集。直接在原始 gaze 上训练，不经过 fixation 过滤，我们的统一方法产生比现有注意力模型更自然的 gaze 轨迹、扫描路径动态和 saliency 映射，为动态环境中人类注意力的时间建模提供有价值的见解。
英文摘要：
Accurately modelling human attention is essential for numerous computer vision applications, particularly in the domain of automotive safety. Existing methods typically collapse gaze into saliency maps or scanpaths, treating gaze dynamics only implicitly. We instead formulate gaze modelling as an autoregressive dynamical system and explicitly unroll raw gaze trajectories over time, conditioned on both gaze history and the evolving environment. Driving scenes are represented as gaze-centric graphs processed by the Affinity Relation Transformer (ART), a heterogeneous graph transformer that models interactions between driver gaze, traffic objects, and road structure. We further introduce the Object Density Network (ODN) to predict next-step gaze distributions, capturing the stochastic and object-centric nature of attentional shifts in complex environments. We also release Focus100, a new dataset of raw gaze data from 30 participants viewing egocentric driving footage. Trained directly on raw gaze, without fixation filtering, our unified approach produces more natural gaze trajectories, scanpath dynamics, and saliency maps than existing attention models, offering valuable insights for the temporal modelling of human attention in dynamic environments.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
