Gaze Influenced by Image Transformations Dataset

Gaze Influenced by Image Transformations Dataset是由上海交通大学图像通信与网络工程研究所创建的一个新型视觉注意力数据集，包含10名观察者在1900张经过19种不同转换处理的图像上的注视点数据。该数据集旨在研究图像转换对人类视觉注意力的影响，并通过分析眼动数据，发现观察者在转换后的图像上注视的位置与原始图像不同。数据集的应用领域包括训练深度视觉注意力模型，以提高模型对非标准刺激的鲁棒性，解决图像转换对视觉注意力预测模型的影响问题。

Gaze Influenced by Image Transformations Dataset is a novel visual attention dataset developed by the Institute of Image Communication and Network Engineering, Shanghai Jiao Tong University. It contains gaze point data collected from 10 observers viewing 1900 images processed with 19 distinct image transformations. This dataset aims to investigate the impact of image transformations on human visual attention. By analyzing eye-tracking data, it reveals that the fixation positions of observers on transformed images differ from those on the original images. Its application scenarios include training deep visual attention models to enhance their robustness against non-standard stimuli, and addressing the issue of how image transformations affect visual attention prediction models.