原书出处
Foundations of Computer Vision (Antonio Torralba, Phillip Isola etc.)
This loss assumes that it is possible to estimate the real depth of each point, which might not be possible in general due to the scale ambiguity. Often, depth may only be estimated up to a global scaling factor. The scale-invariant loss addresses this by eliminating the global scale factor from the calculation:
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/c043.xhtml
EPUB SHA-256:60a7eebc43f5d9a675d776fed3736fbdaa791e55877023f2f8cba016a9b384f8