I have the AI Hat+ 26 TOPS on my Raspberry Pi5. When trying to run the fast-depth precompiled model from model zoo from: hailo_model_zoo/docs/public_models/HAILO8/HAILO8_depth_estimation.rst at master · hailo-ai/hailo_model_zoo · GitHub, I’m not too sure on how to obtain the output from the inference. I’m running “hailo run fast_depth.hef --input-files image.bin” with image in the shape 224,224. It seems to infer successfully and doesn’t return any errors. However, neither does it return any outputs. Does anyone have a clue on how to retrieve the depth map, or the outputs?
Welcome to the Hailo Community!
The HailoRT CLI run tool was designed to test run models to measure FPS, power and PCIe bandwidth and not to run inference for every model. Many models need post processing to make use of the results.
Have a look at the tutorials on the Hailo AI Software Suite docker. They show you also how to run inference on your images for evaluation purpose in the emulator and real hardware. Inside the docker call:
hailo tutorial
We also have application examples in TAPPAS and our Example repository.