thancaocuong

I think they normalize the X and Y coordinates according to the width and height of the image.

@shuxuzhipeng following this example with custorm [yolo](https://github.com/AlexeyAB/deepstream-plugins), we can write the poseplugin to integrate with deepstream.

@twmht I have some suggestions for you: 1. Try lightweight backbone such as mobilenet, ... 2. Use C++

@sunbin1357 can you show me your modifications? I will check it out and give you some suggestions

@ishang3 first you need to follow the deepstream apps, for convinience, I think you need to write inference code in C++. It's like Pose plugin for deepstream. Then write the...

I did implement the preprocess on cuda including convert to RGB, resize, normalize. I will reoganize my code then share with you. Hope it will help

@ishang3 please take a look at nvPreprocess function. I use cuda to normalize RGB image and DMA it to GPU for inference. You also can use my repo for trt_pose,...

hi @IsmaelElHouas, 251 FPS you got is only inference time (without postprocessing time called paf process). Anyway, I obtained overall 7fps when running densenet121 320x320 on jetson tx2. So on...

@jasonakon you need to remove trt_pose completely and reinstall. On jetsonNX, and Jetson TX2 I use torch1.5.0 with jetpack 4.4. So you can try to use lastest jetpack version. I...

@SLyra21 you can use python to export plan file from torch2trt generated .trt file. Then load it using tensorRT c++ API and inference.