Jia Guo
Jia Guo
https://github.com/deepinsight/insightface/tree/master/recognition/_tools_/cpp_align
What do you mean by multiple datasets?
我用resnet可以训到99.73, 是否因为训练、测试集都为长度10，降低了测试的难度？
@piaobuliao 是的 可参考我的项目 https://github.com/deepinsight/insightocr
@Sanster SimpleNet结构是参考的这里: https://github.com/diaomin/crnn-mxnet-chinese-text-recognition
@piaobuliao 我用的就是caffe_ocr数据
You can call `onnxruntime.set_default_logger_severity(3)` to avoid warning.
there's pytorch checkpoint here: https://github.com/deepinsight/insightface/tree/master/model_zoo#22-scrfd
It tells you that some operators will be assigned to be run on CPU, so you should specify the CPU provider in the init option.
The time is too far back, you can simply use some local test sets for comparison.