SID4VAM

SID4VAM是由巴塞罗那自治大学计算机视觉中心和圣地亚哥德孔波斯特拉大学CiTIUS共同创建的一个合成图像数据集，包含230张具有已知显著区域的图像。这些图像通过15种不同的低级特征类型生成，如方向、亮度、颜色和大小等，用于视觉注意建模。数据集分为自由观看和视觉搜索两个任务，旨在评估和改进视觉显著性模型的性能，特别是在处理合成图像时的表现。

SID4VAM is a synthetic image dataset jointly developed by the Computer Vision Center of the Autonomous University of Barcelona and CiTIUS at the University of Santiago de Compostela. It contains 230 images with annotated salient regions. These images are generated using 15 distinct low-level feature types, such as orientation, luminance, color, size and others, and are designed for visual attention modeling research. The dataset is structured around two task paradigms: free viewing and visual search, with the objective of evaluating and improving the performance of visual saliency models, particularly when processing synthetic images.