Did anyone ever succeed in exporting a Detectron2 mask rcnn model via ONNX to TensorRT ? I would like to know what kind of performance this gives.
I also have been getting around 6 fps on x86/GPU with Detectron2, and am wondering if it’s possible to run on the NX or something similar as an edge computer any faster (assuming these boards are actually avaiable again in the near future)
1 Like
This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.