YOLO: Real-Time Object Detection

Re:YOLO: Real-Time Object Detection by amelius on Tuesday April 02, 2019 @02:52PM (#19557384)

Can we also get the orientation of each detected object?

[ Parent | on HN ]

Re:YOLO: Real-Time Object Detection by indutny on Tuesday April 02, 2019 @05:26PM (#19558743)

With some changes - yes. I did this in my experimental project: https://github.com/indutny/resistenz/blob/master/python/mode...

The idea is to add an extra 2 params to the output of each classifier cell. Then do L2 normalization on them ( https://github.com/indutny/resistenz/blob/master/python/mode... ) and treat them as a cosine/sine pair.

The loss in this case would be the Euclidean distance between the actual and predicted pairs, which is equal to "2 * (1 - cos(x-y))".

[ Parent | on HN ]

Re:YOLO: Real-Time Object Detection by razeonex on Tuesday April 02, 2019 @03:18PM (#19557651)

Darknet is a framework for Neural Networks, YOLO is more an algorithm focused on object detection I think it could be relatively easy to perform the detection of an object's detection.

[ Parent | on HN ]

Re:YOLO: Real-Time Object Detection by p1esk on Tuesday April 02, 2019 @03:38PM (#19557821)

Does your training dataset supply that info about each object?

[ Parent | on HN ]

Re:YOLO: Real-Time Object Detection by delaaxe on Tuesday April 02, 2019 @03:19PM (#19557654)

Capsule networks would be better suited for that

[ Parent | on HN ]