Problem creating dataset prior to training

Hello, I’m Bernardo from 2827. This is the first year our team tried to use a machine learning tool to identify our TSE in the field. We took four videos, each in different lighting and positions for the TSE, including frames without the TSE. When trying to form a dataset, the bar that says “Processed frames” doesn’t move at all.

Any way to help us with this problem?

Hi Bernardo,

You’ve found a bug. Thank you.

The bug occurs when you have a video that has exactly 1 frame to be included in the dataset. You have a video where you have checked the Ignore Frame checkbox for all but one frame.

I will fix the bug ASAP. In the meantime, I suggest that you uncheck the Ignore Frame checkbox for at least one more frame so that the video has 2 (or more) frames to be included in the dataset.

-Liz