This repository contains the 9th place (out of 241 teams) solution of the DhakaAI traffic detection competition. We incorporated our evolved hyperparameters by training several configurations of yolov5 models on the competition dataset using the ultralytics YOLOv5 repository. Our approach in the competition is briefly described in this poster. The data preprocessing of yolov5 is same as yolov3 and codes are available here. The training and inference codes are provided here. All code and models are under active development, and are subject to modification.
forked from ultralytics/yolov5
-
Notifications
You must be signed in to change notification settings - Fork 0
YOLOv5 in PyTorch > ONNX > CoreML > TFLite
License
hisham32/DhakaAI-yolov5
Folders and files
Name | Name | Last commit message | Last commit date | |
---|---|---|---|---|
Repository files navigation
About
YOLOv5 in PyTorch > ONNX > CoreML > TFLite
Resources
License
Stars
Watchers
Forks
Packages 0
No packages published
Languages
- Python 96.4%
- Shell 2.5%
- Dockerfile 1.1%