hitnet

HITNET-Stereo-Depth-estimation

Input

(Image from https://vision.middlebury.edu/stereo/data/scenes2003/)

Shape : (1, 6, 640, 480)

Output

Shape : (640, 480, 1)

usage

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 hitnet.py

If you want to specify the input image, put stereo images path after the --left, --right option, respectively.
You can use --savepath option to change the name of the output file to save.

$ python3 hitnet.py --left LEFT_IMAGE_PATH --right RIGHT_IMAGE_PATH --savepath SAVE_IMAGE_PATH

By adding the --video option, you can input the directory of stereo video frames.

$ python3 hitnet.py --video STEREO_DATA_DIR

Reference

ONNX-HITNET-Stereo-Depth-estimation

Framework

TensorFlow

Model Format

ONNX opset=12

Netron

hitnet.onnx.prototxt