(from https://github.com/foamliu/Deep-Image-Matting-PyTorch/tree/master/data/alphamatting/input_lowres)
Ailia input shape: (1, 320, 320, 4) input range: (0,1)
Automatically downloads the onnx and prototxt files on the first run. It is necessary to be connected to the Internet while downloading.
For the sample image,
$ python3 deep-image-matting.py
If you want to specify the input image, put the image path after the --input
and --trimap
option.
You can use --savepath
option to change the name of the output file to save.
$ python3 deep-image-matting.py --input IMAGE_PATH --trimap TRIMAP_PATH --savepath SAVE_IMAGE_PATH
By adding the --video
option, you can input the video.
If you pass 0
as an argument to VIDEO_PATH, you can use the webcam input instead of the video file. The trimap is generated by DeeplabV3.
$ python3 deep-image-matting.py --video VIDEO_PATH
Keras 2.3.1
ONNX opset = 10