invalid bounding boxes error during draining faster rcnn

hi everone i am training faster rcnn during training i faced problem of invalid bounding boxes.i used herper function mentioned in faster rcnn and yolo examples in matlab (which is helperSanitizeBoxes function) but this cannot remove invalid boxes .my code is attached.

Answer 1

[T.Nikhil kumar]

Hello Ahmad,

As per my understanding, you want to train a Faster RCNN network on your own dataset and are facing an invalid bounding box error.

The values of valid bounding boxes are expected to be finite, positive, non-fractional, non-NaN and should be within the image boundary with a positive height and width. All the invalid instances of bounding boxes in the dataset need to be deleted or modified to valid values.

As mentioned in the MathWorks example, the ‘helperSanitizeBoxes’ function should help in checking invalid bounding boxes.

Alternatively, I would suggest you try using a similar implementation of a dataset validation function ‘validateInputData.m’ in the ‘Object Detection using YOLOv3 network’ example attached below, for your purpose. You can validate your combined datastore ‘cds’ using this function and remove/modify invalid entries.

Open the example using ‘Copy Command’ and view the definition of the function in the folder.

https://www.mathworks.com/help/vision/ug/object-detection-using-yolo-v3-deep-learning.html

Hope this helps!