关于深度图的问题

[hhqweasd]

你好，深度图确实有用吗？我看消融实验加了深度图只高了0.06分？能提供 得分 vs. 迭代 曲线吗？想入坑这个方向，发现这个方向的问题应该是运动估计的问题，你有什么建议吗？

[baowenbo]

It does work, I didn't try too hard to get the cheerful results.
In Table 4, you can see that my results is about 0.42dB better on Vimeo90K when comparing to my MEMC-Net method that does not take use of depth.

[coolKeen]

It does work, I didn't try too hard to get the cheerful results.
In Table 4, you can see that my results is about 0.42dB better on Vimeo90K when comparing to my MEMC-Net method that does not take use of depth.

Could you please publish a version of DAIN without depth information?