标题： GazeTarget360：面向机器人感知的360度注视目标估计 显示英文标题

标题： GazeTarget360: Towards Gaze Target Estimation in 360-Degree for Robot Perception

摘要： 使机器人理解人类的注视目标是实现下游任务能力的关键步骤，例如在现实世界的人机交互中进行注意力估计和运动预期。 以往的工作通过数据驱动的方法仔细去除帧外样本，解决了帧内目标定位问题。 基于视觉的注视估计方法，如OpenFace，无法有效吸收图像中的背景信息，并且在被试者看向摄像头以外的方向时无法预测注视目标。 在本工作中，我们提出了一种系统，以解决从一般视觉场景图像中进行360度注视目标估计的问题。 该系统名为GazeTarget360，集成了眼接触检测器、预训练的视觉编码器和多尺度融合解码器的条件推理引擎。 交叉验证结果表明，GazeTarget360可以在未见过的场景中生成准确且可靠的注视目标预测。 这使得该系统成为首个能够从真实摄像机画面中预测注视目标的系统，其高效且可部署。 我们的源代码已公开在：https://github.com/zdai257/DisengageNet。 显示英文摘要

摘要： Enabling robots to understand human gaze target is a crucial step to allow capabilities in downstream tasks, for example, attention estimation and movement anticipation in real-world human-robot interactions. Prior works have addressed the in-frame target localization problem with data-driven approaches by carefully removing out-of-frame samples. Vision-based gaze estimation methods, such as OpenFace, do not effectively absorb background information in images and cannot predict gaze target in situations where subjects look away from the camera. In this work, we propose a system to address the problem of 360-degree gaze target estimation from an image in generalized visual scenes. The system, named GazeTarget360, integrates conditional inference engines of an eye-contact detector, a pre-trained vision encoder, and a multi-scale-fusion decoder. Cross validation results show that GazeTarget360 can produce accurate and reliable gaze target predictions in unseen scenarios. This makes a first-of-its-kind system to predict gaze targets from realistic camera footage which is highly efficient and deployable. Our source code is made publicly available at: https://github.com/zdai257/DisengageNet.