现有的显着性检测方法使用图像作为输入，并且对前景/背景相似性，复杂背景纹理和遮挡敏感。我们探讨了使用光场作为显着性检测的输入的问题。(Existing saliency detection approaches use images as in-puts and are sensitive to foreground/background similari-ties, complex background textures, and occlusions. We ex-plore the problem of using light fields as input for saliency detection.)