You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Hi;
I've tried to train pose estimation on MARS with previously annotated csv file and raw images to compare the performance metrics of the model with other pose estimation models such as deeplabcut and sleap. And actually, the original dataset is from MARS itself, which is 15,000 frames from top-taken videos on mice. I used the dataset on deeplabcut then as deeplabcut annotation on sleap and now want to train also on MARS wth the same specifications but getting error which is most probably because of the format mismatch of csv file. I would like to ask, what should be the example csv annotation file if we want to train MARS without labeling, just by previously annotated data?
The text was updated successfully, but these errors were encountered:
Hi;
I've tried to train pose estimation on MARS with previously annotated csv file and raw images to compare the performance metrics of the model with other pose estimation models such as deeplabcut and sleap. And actually, the original dataset is from MARS itself, which is 15,000 frames from top-taken videos on mice. I used the dataset on deeplabcut then as deeplabcut annotation on sleap and now want to train also on MARS wth the same specifications but getting error which is most probably because of the format mismatch of csv file. I would like to ask, what should be the example csv annotation file if we want to train MARS without labeling, just by previously annotated data?
The text was updated successfully, but these errors were encountered: