Alvin

Can you provide C + + face alignment code

2

i can't fin https://github.com/deepinsight/insightface/blob/master/recognition/tools/cpp-align/FacePreprocess.h

@marcoslucianops Thanks for your work Looking forward to updating Dynamic batch-size

Using PyTorch and TensorRT, the results were quite different

4

I trained 2 classes of behavior recognition using MobilenetV2 and then used Torch and TensorRT on the Jetson Nano to get vastly different results. Excuse me, what's the problem? model...

Why data enhancement on test datasets,not on train datasets?

1

https://github.com/jfzhang95/pytorch-video-recognition/blob/ca37de9f69a961f22a821c157e9ccf47a601904d/dataloaders/dataset.py#L81

感谢大佬分享~ demo 我已经跑通，不过我使用多进程运行的时候，遇到了以下错误。 [TensorRT] ERROR: ../rtSafe/cuda/reformat.cu (925) - Cuda Error in NCHWToNCHHW2: 400 (invalid resource handle) [TensorRT] ERROR: FAILED_EXECUTION: std::exception 希望大佬指教。感谢！

Hi~ It does not work when executing Python programs (playing audio with playsound) through Supervisor Other functions are normal except playsound does not work my config like this: [program:ocr] directory=/home/firefly/worspace/PPOCR...

How to train next

1

model: mobileface device: NVIDIA 2080 * 1 dataSet: glint360K Based on the pre training model, continue training for about 30 days loss is reduced only in a few iterations after...

@TreB1eN In this case, is the head.kernel parameter updated?

about hyper parameter settings in my own dataset

1

hi, @TreB1eN Thank you for sharing. I set up a dataset from YouTube video, AFDB, CNBC etc, with a total of 6587 categories. Use MTCNN to detect face and align...