i am running into the following issue when running either sample_object_tracker, sample_object_detector or sample_object_dwdetector on Drive PX2 with the latest firmware flashed:
nvidia@nvidia:/usr/local/driveworks-0.3/bin$ ./sample_object_detector Program Arguments: --tensorRT_model=.././data/samples/detector/tensorRT_model.bin --video=.././data/samples/sfm/triangulation/video_0.h264 --yuv2rgb=cuda Initialize DriveWorks SDK v0.3.400 Release build with GNU 4.9.2 from v0.3.0-rc8-0-g3eeebea against PDK v220.127.116.11 SDK: Resources mounted from /usr/local/driveworks-0.3/data/resources SDK: Create NvMediaDevice SDK: use EGL display as provided SDK: dGPU has been detected as device number 0 SDK: iGPU has been detected as device number 1 SDK: Default GPU Device set to device number 0 SAL: identified board as DrivePX2-TegraA SensorFactory::createSensor() -> camera.virtual, video=.././data/samples/sfm/triangulation/video_0.h264,yuv2rgb=cuda CameraNvMedia: no seek table found at .././data/samples/sfm/triangulation/video_0.h264.seek, seeking is not available. Camera image with 1280x800 at 30 FPS Initializing TensorRT Network: .././data/samples/detector/tensorRT_model.bin Driveworks exception thrown: DW_CUDA_ERROR: Couldn't create cudnn Cannot initialize TensorRT Network: DW_CUDA_ERROR
The windows remains open and displays the video without any detection.
I also cross compiled the sample from scratch using the instructions provided and still get the same error. I made sure that the tensorRT_model.bin exists in the desired path.
Any help is greatly appreciated.