Is there a training function for yolov3(tra​inYOLOv3Ob​jectDetect​or)? similar to trainYOLOv​2ObjectDet​ector(yolo v2) and trainYOLOv​4ObjectDet​ector(yolo v4)

Please let me know if there is any training function for YOLO v3. As i could see in the example, YOLOv3 it does the training using minibatchqueue, but the examples of YOLO v2 and v4 both uses a training function trainYOLOv2ObjectDetector and trainYOLOv4ObjectDetector respectively.

If not, how can i give trainingoptions as input during training?

Thank you in advance.

[Walter Roberson]

https://www.mathworks.com/help/vision/ug/object-detection-using-yolo-v3-deep-learning.html shows an example of training YOLO3 but there does not appear to be a specific function.

Answer 1

[Neha]

I understand that you want to specify the hyperparameters through a YOLO v3 training function, but since this is not possible in YOLO v3, Object Detection Using YOLO v3 demonstrates a custom training loop.

The variables for the hyperparameters are initially defined here: Training Options in YOLO v3

miniBatchSize hyperparameter has been used in the "minibatchqueue" function while the rest of the variables for the hyperparameters have been used in the custom training loop.