Egor Burkov

I personally don't :( unfortunately09:04, May 29, 2023, "小王同学" ***@***.***>: Does anyone have trained X2Face meta model? —Reply to this email directly, view it on GitHub, or unsubscribe.You are receiving...

Looks like you caught a bug. These heatmaps seem unnatural to me, as if the network is underfit or overfit. For example, I believe there shouldn't be such checkerboard patterns....

Hi, unfortunately I don't have logs :( I still bet you have a bug somewhere in the backbone. I suggest that you try checking activations and gradients layer-wise.

Do you really have `/usr/local/opt/ffmpeg/lib/libavcodec*.dylib`? If yes, [this link](http://stackoverflow.com/questions/14358950/updating-a-dylib-to-reference-correct-versions-of-other-dylibs-homebrew) suggests reinstalling ffmpeg then opencv. If it doesn't work, try `DYLD_FALLBACK_LIBRARY_PATH` trick as shown [here](http://stackoverflow.com/questions/37949126/dyld-library-not-loaded-reason-image-not-found-mac-osx-el-capitan).

Add something like `[:100]` [here](https://github.com/karfly/learnable-triangulation-pytorch/blob/9d1a26ea893a513bdff55f30ecbfd2ca8217bf5d/train.py#L98)

Looks like the package responsible for loading Caffe nets fails on Jetsons: https://github.com/szagoruyko/loadcaffe/issues/35 I don't have one so can't reproduce :(

Hi, we don't know. Better ask the original authors. By the way, Karim extracted [camera parameters](https://github.com/karfly/human36m-camera-parameters) (but not keypoints).

That's strange, are you on a fresh version of (Torch and) LuaJIT?

You can insert an extra "person" into the gallery called "Unknown" and assign this label to all the unknown people. Or, if you want to experiment with the code, you...

Sorry for delay.You may get the wrong thing with SVM's "confidences" (which are not really a natural thing for it). I suggest 2 alternatives:\* Read [this research on open set...