Opencv dnn doesnot use GPU
Mamta-B opened this issue · 2 comments
Mamta-B commented
I m using the following to detect objects using SSD mobilenet (CAffe) on Jetson TX2 with opencv3.4.1 and Cuda 9.0
blob = cv2.dnn.blobFromImage(cv2.resize(frame, (300, 300)), 0.007843, (300, 300), 127.5)
net.setInput(blob)
detections = net.forward()
The tegrastats show no GPU usage at all. All the cores of the CPU are utilized properly, however GPU utilisation is zero.
How can I enable GPU usage on the Jetson for faster FPS on videos ?
Thanks
jetsonhacks commented
Are you using Python?
jetsonhacks commented
Are you linked against the OpenCV 3.4 libraries in Caffe? Without providing more information, it is difficult to diagnose your issue.