YOLOv5 is a family of object detection architectures and models pretrained on the COCO dataset, and represents Ultralytics
open-source research into future vision AI methods, incorporating lessons learned and best practices evolved over thousands of hours of research and development.
This is a customised version for corn plant identification as part of an ongoing research project.
cornmodem864i12002
Colab trained model with 1,200 images (First50 dataset) containing 2,727 class instances
--img 864 --batch 8 --epochs 150 --cfg models/yolov5l.yaml
Peak performance in 23rd epoch
[email protected]:
kornmodel960m
Locally trained model with 684 images
--img 960 batch 4 --epochs 100 --data korn.yaml --cfg models/yolov5m.yaml
Computing time: ~31 hours
[email protected]: unknown
kornmodel960l_1200
Colab trained model with 1200 images
--img 960 --batch 8 --epochs 150 --data korn.yaml --cfg models/yolov5l.yaml
Computing time: ?
[email protected]: unknown
Labelled with labelImg
- slice_images.py for resizing HR input images
- split_train_val.py for dividing data and structuring into Yolo-readable format
- count_instances.py to count class instances in labels
- find_kornrows.py to find the orientation of cornrows in the field