I converted dashcamnet into an engine (.trt) file, and i am trying to run it directly with TensorRT c++ api. The forward pass run successfully, However I cannot find any information on how to convert theses normalized boxes into absolute ones. It would be great if you guys can show me the equation or code snippet. Thank you.
For postprocessing of dashcamnet, please refer to /opt/nvidia/deepstream/deepstream/sources/libs/nvdsinfer_customparser/nvdsinfer_custombboxparser.cpp
1 Like