原书出处
Foundations of Computer Vision (Antonio Torralba, Phillip Isola etc.)
These models can be easily understood by first considering the problem of synthesizing one pixel, then two, and so on. The first observation to make is that it's pretty easy to synthesize a single grayscale pixel. Such a pixel can take on 256 possible values (for a standard 8-bit grayscale image). So it suffices to use a categorical distribution to represent the probability that the
原书定位详情
来源:a19c7933-4eda-4176-aebd-30fe7e1ad292
正文位置:extracted/Foundations of Computer Vision (Antonio Torralba, Phillip Isola etc.)/_epub-src/OPS/xhtml/c032.xhtml
EPUB SHA-256:60a7eebc43f5d9a675d776fed3736fbdaa791e55877023f2f8cba016a9b384f8