xipeng13 / recurrent-face-alignment Goto Github PK

when i run this code ,i get this error message:
Message type "caffe.LayerParameter" has no field named "upsample_param".

Impressive works!
Is there demonstration link on Youtube or else?

Hi,
When I run this code, I get some error like below:

Error parsing text-format caffe.NetParameter: 48:12: Message type "caffe.LayerParameter" has no field named "bn_param".

Could you help me?

Hello. Thanks your impressive works!
Which face detector method do you use when processing image?
And where can I find your train/test codes?

Hello Xipeng,

I am impressed by your work using RNN and encoder decoder network to solve the sequential face alignment problem. Really nice job!

I would like to know if your code is compatible with officially released caffe version since it supports batch normalization as well. [https://github.com/BVLC/caffe/blob/master/include/caffe/layers/batch_norm_layer.hpp](BVLC-caffe-batch norm)

I found that the bn_param in your deploy is different from the official version and might need your modified version of caffe.

Thank u in advance and wish you all the best.

Hello, I am trying to run your code, but caffe is giving me errors like: "Message type "caffe.LayerParameter" has no field named "bn_param"."
I am using the latest version from the oficial repository.

So, which version of the caffe are you using?

Thank you.

In the research paper, I noticed that this model was compared with others on 68 landmarks. If possible, please can you upload the pre-trained weights on 68 landmarks. Also, if possible can you add code to train the model that would be very helpful.

Hi, I found a error in the code. The Lmk2Bbox_7lmks was in the PXTrack.py not in the PXPts.py.
But in all *Demo.py, you use the PXPTs.Lmk2Bbox_7lmks to call the function. It's a fault.

[libprotobuf ERROR google/protobuf/text_format.cc:298] Error parsing text-format caffe.NetParameter: 49:22: Message type "caffe.BatchNormParameter" has no field named "scale_filler".
F0514 18:03:26.403556 2532959168 upgrade_proto.cpp:88] Check failed: ReadProtoFromTextFile(param_file, param) Failed to parse NetParameter file: /Users/cogsys/Downloads/recurrent-face-alignment-master/model/deploy.prototxt

I used "https://github.com/TimoSaemann/caffe-segnet-cudnn5" and succeed to compile. Also rpath is set properly.

Have no idea about "scale_filter" issue...