Multiple types of NN model optimization environments. It is possible to directly access the host PC GUI and the camera to verify the operation.
- Docker 20.10.5, build 55c4c88
- Ubuntu 18.04 x86_64
- TensorFlow v2.4.1 (MediaPipe Custom OP, FlexDelegate, XNNPACK enabled)
- tflite_runtime v2.4.1 (MediaPipe Custom OP, FlexDelegate, XNNPACK enabled)
- edgetpu-compiler
- flatc 1.12.0
- TensorRT cuda11.0-trt7.1.3.4-ga-20200617
- PyTorch 1.7.1+cu110
- TorchVision 0.8.2+cu110
- TorchAudio 0.7.2
- OpenVINO 2021.3.394
- tensorflowjs
- coremltools
- onnx
- tf2onnx
- tensorflow-datasets
- openvino2tensorflow
- tflite2tensorflow
- onnxruntime
- onnx-simplifier
- gdown
- OpenCV 4.5.2-openvino
$ xhost +local: && \
docker run -it --rm \
--gpus all \
-v `pwd`:/home/user/workdir \
-v /tmp/.X11-unix/:/tmp/.X11-unix:rw \
--device /dev/video0:/dev/video0:mwr \
-e DISPLAY=$DISPLAY \
--privileged \
pinto0309/mtomo:ubuntu1804_tf2.4.1_torch1.7.1_openvino2021.3.394
$ git clone https://github.com/PINTO0309/mtomo.git && cd mtomo
$ docker build -t {IMAGE_NAME}:{TAG} .
$ xhost +local: && \
docker run -it --rm \
--gpus all \
-v `pwd`:/home/user/workdir \
-v /tmp/.X11-unix/:/tmp/.X11-unix:rw \
--device /dev/video0:/dev/video0:mwr \
-e DISPLAY=$DISPLAY \
--privileged \
{IMAGE_NAME}:{TAG}