a wide range of pixel-wise tasks without structural adaptation
    semantic segmentation，real image denoising，portrait image matting，night image enhancement
    sparkle：
    （1）flaw detector来求置信图（pixel-wise）
    （2）不依赖特定任务
    Limitations：
    F只受labeled训练，不稳定，CPS解决

    Contribution:
    (1) address the issues caused by the dense outputs through a novel flaw detector.
    (2) learn from unlabeled data collaboratively through two newly proposed constraints that are independent of task-specific properties.
    image.png

    image.png
    task models：
    image.png
    image.png
    image.png
    image.png
    Flaw Detector ：
    image.png
    image.png
    discriminator vs flaw detector
    都是来approximate the prediction confidence
    （1）the flaw detector predicts a dense probability map with location information while the discriminator predicts an image-level probability.
    （2） we use the ground truth of the labeled data to generate the targets of the flaw detector.

    就是说： D averages all predicted pixels to get a single confidence value during training, as its target is an image-level real or fake probability. Using an average confidence to represent the overall confidence is not appropriate.