Openvino2tensorflow

- CUDA 11.6
- TensorRT 8.4.0
- cuDNN 8.4
- TensorFlow v2.9.0
- TensorFlow Lite v2.9.0
- PyTorch v1.12.0 (with grid_sample)
- onnxruntime-gpu v1.12.0

- https://github.com/PINTO0309/openvino2tensorflow/pull/106 transpose channel for prelu replacement when prelu channel number is 3 zye1996
- https://github.com/PINTO0309/simple-onnx-processing-tools v1.0.22

Added workaround for avoiding conversion errors.
- GroupConvolution
- MatMul
- NonZero

- CUDA 11.6
- TensorRT 8.4.0
- cuDNN 8.4
- TensorFlow v2.9.0-rc0
- TensorFlow Lite v2.9.0-rc0
- PyTorch v1.12.0 (with grid_sample)
- onnxruntime v1.12.0

- Applying NCHW<->NHWC transposition to **`Subtract`** 4D constant
- OpenVINO 2022.1 Added logic to ignore the second output of **`MaxPool`** with opset=8
- https://github.com/PINTO0309/openvino2tensorflow/issues/103 openvino int8/fp32 to tflite convertion
- https://github.com/PINTO0309/openvino2tensorflow/issues/100 KeyError
- https://github.com/PINTO0309/openvino2tensorflow/issues/99 KeyError when running openvino2tensorflow

https://github.com/PINTO0309/simple-onnx-processing-tools
1. sne4onnx
2. snd4onnx
3. snc4onnx
4. scs4onnx
5. sog4onnx
6. sam4onnx
7. soc4onnx
8. scc4onnx