Tracking with Efficient Re-Identification in Ultralytics
Add reidentification to Ultralytics without compromising latency using object-level features extracted from YOLO.
Hello. I'm Mohammed Yasin.
I develop computer vision solutions for video analytics at the edge.
Use Lightly to pretrain a YOLO backbone through self-supervised learning and then fine-tune it in Ultralytics. Boost your inference performance by exploiting Kalman filter predictions to interpolate detections for skipped frames. Extract the object-level features from YOLO for downstream tasks without extra overhead. Implement class balancing in Ultralytics using a weighted dataloader and improve the performance of minority class.
Recent Posts
Pretrain YOLO Backbone Using Self-Supervised Learning With Lightly
Boosting Inference FPS With Tracker Interpolated Detections
Retrieving Object-Level Features From YOLO
More Posts
Balance Classes During YOLO Training Using a Weighted Dataloader
My Projects
View all »