Hello everyone,
I am currently working on a graduation project that involves building a functional autonomous driving vehicle. One of it's features is to use OAK-D camera to detect objects and calculate the distance between it and the camera.
Right now I have been able to run spatial detection from this source but using a YOLO-v8s blob/config
https://github.com/luxonis/depthai-python/blob/main/examples/SpatialDetection/spatial_tiny_yolo.py
My question is, how can I use a custom YOLO-v8 dataset with my own classes that is relative to a driving environment?
This decision was reached because we believe that by cutting down the classes and leaving the necessary ones could prove to be efficient in terms of performance. Yolo-v8 gives us 10fps or less which is a performance far from what we need.
Thanks!
