DeepStream Faster-RCNN model

I am a beginner of DeepStream.

GIE can implement Faster-RCNN(VGG) model?

Hi Hnwyshk,
In Deepstream, we can’t suppory Faster-RCNN(VGG) because some layer can’t be supported in TensorRT.

For evaluating network, you can use TensorRT directly.

Thanks
wayne zhu

Thank you for your answer.

For example, is it possible to implement FasterR-CNN+TensorRT with reference to the dsexample plugin?

Thanks.

Yes,
In dsexample, you can get raw YUV buffer, so it is ok for you to add tensorRT code.

Thanks
wayne zhu