MOS-Multi-Task-Face-Detect

MOS-Multi-Task-Face-Detect copied to clipboard

TensorRT ?

1

TensorRT implementation ?

dexception

I let device = torch.device('cpu') in detect_video.py, then the speed drop dramatically from 60fps to 2fps.

20XYSong

Question about 300W-LP training

1

Thanks again for your work. Can you please provide the training details of the the MOS-S-300wlp model? How did you define the supervision, and what are your training settings?

FunkyKoki

Thanks again for your work. Since you use flip and multi-scale strategy in testing, I wonder how your model performs when these test tricks are not applied. I ask this...

FunkyKoki

论文中说以半监督的形式标注了不小于 35×35 像素的人脸，但是我发现公开的数据集中小于 35×35 像素的人脸也有姿态信息，请问这些姿态信息可靠吗？

QAQEthan

Hi expert, I try to verify histogram with your "label.txt" on Figure 3(b) but I found large difference between Figure 3(b) and label.txt. The following statistics from "label.txt": data size:...

BossunWang

train code

2

thanks for your great work.but i did not find some code for train. Do you plan to open the train code?

abcliguanxi

Convert to ONNX

1

I have a question about ONNX format. Can you convert MOS.pth to MOS.onnx?

huylgia

It may be a DCNv2 error, i use cuda10.1

FrankieMAN-YU