This project is based on ultralytics/yolov3.

LF-YOLO (Lighter and Faster YOLO) is used to detect defect of X-ray weld image.

Download

$ git clone https://github.com/lmomoy/LF-YOLO

Train

We provide multiple versions of LF-YOLO with different depths.

$ python train.py --data coco.yaml --cfg LF-YOLO.yaml      --weights '' --batch-size 1
                                         LF-YOLO-1.25.yaml                           1
                                         LF-YOLO-0.75.yaml                           1
                                         LF-YOLO-0.5.yaml                            1

Results

We test our model on public dataset MS COCO, and it also achieves competitive results.

Model size
(pixels)		 mAPval
0.5:0.95		 mAPtest
0.5:0.95		 mAPval
0.5		 params
(M)		 FLOPS
640 (B)
YOLOv3-tiny 640 17.6 17.6 34.8 8.8 13.2
YOLOv3 640 43.3 43.3 63.0 61.9 156.3
YOLOv3-SPP 640 44.3 44.3 64.6 63.0 157.1
LF-YOLO 640 27.8 27.9 47.8 7.4 17.1
Table Notes (click to expand)
  • APtest denotes COCO test-dev2017 server results, all other AP results denote val2017 accuracy.
  • AP values are for single-model single-scale unless otherwise noted. Reproduce mAP by python test.py --data coco.yaml --img 640 --conf 0.001 --iou 0.65
  • SpeedGPU averaged over 5000 COCO val2017 images using a GCP n1-standard-16 V100 instance, and includes FP16 inference, postprocessing and NMS. Reproduce speed by python test.py --data coco.yaml --img 640 --conf 0.25 --iou 0.45
  • All checkpoints are trained to 300 epochs with default settings and hyperparameters (no autoaugmentation).

Requirements

Python 3.8 or later with all requirements.txt dependencies installed, including torch>=1.7. To install run:

$ pip install -r requirements.txt

Inference

$ python detect.py --source data/images --weights LF-YOLO.pt --conf 0.25

GitHub

View Github