Hello Team,
I have compiled face detection model using edgeai-tidl-tools (version 11_01_06_00). The inference script on ONNX runtime is working fine.
As a next step, I need to run this compiled model artifacts in Vision Apps OpenVx pipeline. I am following /J784S4/ti-processor-sdk-rtos-j784s4-evm-11_01_00_04/vision_apps/apps/dl_demos/app_tidl/ as an example, did the required pre-processing, but I am not getting proper results after processing the graph.
Is there any thing that is different here that needs to be done with respect to preprocessing/postprocessing for making it work in OpenVx pipeline?
Best Regards,
Reshma