dexception
dexception
Like the first picture but 2 detections for each object. One horizontal and one vertical thus making a plus sign. Out of 20 objects 12 are having this plus sign.
I am able to run it on ubuntu 16.04, cuda 9.0, cudnn 7.0.5, python2
@wondonghyeon Is this a regular tar file that can be extracted or we just need to use it as it is ?
Traceback (most recent call last): File "pred.py", line 118, in main() File "pred.py", line 72, in main model.load_state_dict(torch.load(f)['state_dict']) File "/usr/local/lib/python3.5/dist-packages/torch/serialization.py", line 387, in load return _load(f, map_location, pickle_module, **pickle_load_args) File...
Got it ... This works with python2 and not python3
@Bkmz21 Thanks for replying ! ``` root@xyz:/home/xyz# ./CNNObjectDetector TEST ACCURACY [CNNDetector] Initializing with max image size (1280, 720)! [TEST ACCURACY] test allocate memory [TEST ACCURACY] cannels = 1: success [TEST...
@Bkmz21 Disabled Image Converter, Resizer, CNN [All of them are giving segmentation fault] Changed BUILD_FaceDetector to ON inside CMakeLists.txt but there is not binary to test it. Please review the...
@Bkmz21 Tried it on windows. Not accurate as claimed in the paper.