Xuan Wang comments

Results 12 comments of


                                            Xuan Wang

RuntimeError: size of dimension does not match previous size, operand 1, dim 2

> yes！my problem was solved after I changed the smpl model. (the right one is smpl_v1.0.0)

Compiling mediapipe/gpu/gl_texture_buffer.cc failed: (Exit 1): crosstool_wrapper_driver_is_not_gcc failed: error executing command external/local_config_cuda/crosstool/clang/bin/crosstool_wrapper_driver_is_not_gcc -MD -MF bazel-out/k8-opt/bin/mediapipe/gpu/_objs/gl_texture_buffer/gl_texture_buffer.d

Well, the error still occur after I updating the gcc version to 9.4.0. Is there any missing procedures ?

How to generate the ears and necks in the demo？Do I need to retrain the model？

> > Hi, if you want to generate ears and necks, you need to use the full BFM09 face model. We excluded ear and neck region of original BFM09 to...

I have now trained the model, I want to do model evaluation, can you share your model evaluation code. Much appreciated.

Same problem. Thanks for your attention!

Wrong landmarks of full region

I think it may be causded by **mean_shape** , but I am not sure and cannot solve it.

Wrong landmarks of full region

> Hi, I get the same output results for all the datasets from examples that provided by author, may I know do you face the same problem? I‘m very sorry...

Wrong landmarks of full region

> > > Hi, I get the same output results for all the datasets from examples that provided by author, may I know do you face the same problem? >...

Wrong landmarks of full region

> > > > > Hi, I get the same output results for all the datasets from examples that provided by author, may I know do you face the same...

Wrong landmarks of full region

@lingleong981130 hi~ I'm not sure exactly what errors you mentioned. When I trained with a sub-dataset, about 15000 photos, I only met the problem of "tensorflow" version, which is used...

Evaluation Code

Hello~I think that I really need the **evaluation code** to get the error metric. Donot know if you have any solution now？ Thank you for your attention！