CBSR_WACV_CODE
CBSR_WACV_CODE
> Conv_165 Conv BatchNormalization_166 BatchNormalization Shape_167 Shape Traceback (most recent call last): File "convertCaffe.py", line 122, in convertToCaffe(graph, opset_version, prototxt_path, caffemodel_path) File "convertCaffe.py", line 64, in convertToCaffe err.unsupported_op(node) File "/home/wyw/ONNXToCaffe/onnx2caffe/_error_utils.py",...
嗷，可能是预处理这里不太一致，我这里normalize了，您的test代码是没有normalize的对吧？
如果训练的时候normalize了，测试不normalize应该差距会很大吧？毕竟不是一种输入了？
好的，谢谢您，我再测试下看看，新年快乐
When I try to convert the IR50 onnx to caffe, there is an error about PReLU
I wonder what's the datasets for masked face recognition testing?
> 你可以考虑使用insightface中的partial_fc方法，这个可以解决你的问题。但是需要该一定的代码，我已经将adaface添加到insightface中并成功解决了你的这个问题 请问您是怎么合并partial fc的？可以请教一下吗？
The results are still diverse for each process and the results are different from setting capacity_factor=1.25.
Sorry to bother you again， and I'm wondering how to test a model in MFR-21 if I did not participate in this competition?
请问您是否在MFR2上测试了？