[Gen2] FastDepth on DepthAI

This example shows an implementation of FastDepth on DepthAI in the Gen2 API system. Blob is created with ONNX from PINTO's Model ZOO, which is then converted to OpenVINO IR with required flags and converted to blob.

There are two available blob's for different input sizes:

• fast_depth_256x320_openvino_2021.4_6shave, for 256x320 (H x W) input, which runs with ~30 FPS, and
• fast_depth_480x640_openvino_2021.4_6shave, for 480x640 (H x W) input, which runs with ~10 FPS.

Pre-requisites

Install requirements:

python3 -m pip install -r requirements.txt

Usage

python3 main.py [options]

Options:

• -w, --width: Select width of the inference model. Default: 320. Possible: 320 or 640.