AlexeyAB / darknet

YOLOv4 / Scaled-YOLOv4 / YOLO - Neural Networks for Object Detection (Windows and Linux version of Darknet )
http://pjreddie.com/darknet/
Other
21.79k stars 7.97k forks source link

How to use the new tracking features #6215

Open haviduck opened 4 years ago

haviduck commented 4 years ago

Hello, As im working on a multi-object tracking feature for yolo and python, the new commit caught my eye with the tracking feature. Anyone have any info on what it does and how to use it? ive scoured the code but cant find anything that i can call from py.

Im on a yolo4-custom model, python 3.7, cuda 10.1. Im calling darknet detection pretty much just like the example does it. Debug is throwing some info, but what im really after is a potential unique track_id.

flowzen1337 commented 4 years ago

I'm also interested in it with exactly the same environment, python 3, cuda 10, custom yolov4 dataset and python.

haviduck commented 4 years ago

i had an older repo, the relevant cpp code is there now :) looks like i gotta read up on the New layers

AlexeyAB commented 4 years ago

This is in progress https://github.com/AlexeyAB/darknet/issues/6004

haviduck commented 4 years ago

This is in progress #6004

sweet! the cpp example is crazy cool.