A repository for storing models that have been inter-converted between various frameworks. Supported frameworks are TensorFlow, PyTorch, ONNX, OpenVINO, TFJS, TFTRT, TensorFlowLite (Float32/16/INT8), EdgeTPU, CoreML.
In conversion from Openvino IR (.xml + .bin) to tflite /onnx the order of inputs to the last Concat where the rest of the network is concatenated with "head_pose_angle" is switched. It produces the wrong result. Is there a way to switch the order?
Issue Type
Bug
OS
Ubuntu
OS architecture
x86_64
Programming Language
Python
Framework
ONNX
Model name and Weights/Checkpoints URL
091_gaze-estimation-adas-0002 network (https://github.com/PINTO0309/PINTO_model_zoo/tree/main/091_gaze-estimation-adas-0002)
Description
In conversion from Openvino IR (.xml + .bin) to tflite /onnx the order of inputs to the last Concat where the rest of the network is concatenated with "head_pose_angle" is switched. It produces the wrong result. Is there a way to switch the order?
Relevant Log Output
URL or source code for simple inference testing code
https://github.com/openvinotoolkit/open_model_zoo/tree/2021.2/demos/gaze_estimation_demo