Ali Darwish, Ph.D.

icon indicating an email [email protected]

icon company Microsoft icon location PaloAlto, CA icon location Senior Data Scientist at Microsoft. Multi AI agent reasoning researcher. Robotics software and perception engineer.

Results 45 comments of Ali Darwish, Ph.D.

Key discriminator/d_bn0/beta not found in checkpoint?

Same issue here.

initial bias towards action=1?

Same issue here as @wobeert described. It didn't change even after 622000 steps. See below ![image](https://user-images.githubusercontent.com/10216145/42135614-2e2830ec-7d1b-11e8-9263-8a58300f73e6.png)

initial bias towards action=1?

I fixed it! I introduced a bug by mistake to the original code when I was creating multigpu version for gpu keras. Thanks!

Issue compiling on Ubuntu and tensorflow 1.4

Changing "-D_GLIBCXX_USE_CXX11_ABI=0" to "-D_GLIBCXX_USE_CXX11_ABI=1" worked for me.

Error while executing DDQN for defend_the_center.cfg

Same error here, have you fixed it @yculcarnee ?

Error while executing DDQN for defend_the_center.cfg

The problem is that the passed list has two elements only. commenting out that line should work for now.

Any plans to support OpenCV > 3

Check out this repository , it has cv2 v3 support https://github.com/toinsson/CMT All what you have to do is this if cv2.__version__ < 3.0: self.detector = cv2.FeatureDetector_create(self.DETECTOR) self.descriptor = cv2.DescriptorExtractor_create(self.DESCRIPTOR) self.matcher...

add support for opencv 3+

Awesome!

Segmentation fault (core dumped)

Thank you for your quick response and for helping out. This is all what I'm getting Segmentation fault (core dumped) I've tested it in python 2 and 3, didn't work.

Retraining SSD with Udacity's self driving car dataset

Hi Peeranat, I'll try to answer your question base on the little information that you have provided. Lets start with the performance that you see when you test your classifier...