Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Research to follow : Tracker "baked in" YOLO #318

Open
tdurand opened this issue Nov 25, 2020 · 3 comments
Open

Research to follow : Tracker "baked in" YOLO #318

tdurand opened this issue Nov 25, 2020 · 3 comments

Comments

@tdurand
Copy link
Member

tdurand commented Nov 25, 2020

AlexeyAB/darknet#6004

Sounds very interesting as current state of the art tracking method is DeepSORT which use features from the images to perform tracking.. but is very slow cause of this..

From what I understand there.. YOLO already mainpulate / extracts those images features when performing object detection and the idea behind YOLOv4-tiny-contrastive is to also use this data to perform tracking without impacting FPS / performance.

If this works... I guess we will be able to say goodbye to our simple https://github.com/opendatacam/node-moving-things-tracker ;-)

@rantgithub
Copy link

I understood the same.

Yes, if this works there wont be needed the tracker because now is already built in in Yolo

@haviduck
Copy link

im using a modified version of sort, its predecessor. real easy to use and no tf deps

@rantgithub
Copy link

Hi

just to add the sample code mentioned on the darknet doc f you havent seen this already for detection and tracking

https://github.com/AlexeyAB/darknet/blob/master/src/yolo_console_dll.cpp

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Projects
None yet
Development

No branches or pull requests

3 participants