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
Thanks for all of your hard work.
I have created a custom dataset from scratch using sequence to sequence image frames from a video.
I have spent countless hours manually annotating all of the frames one by one, creating bounding boxes around each specified object and gave them their classification names.
I then saved the dataset as VOC file seeing that the images will be in jpg format and the annotation details will be in XML format.
Now here's where the issues occur, once I have clone your repo, I then upload the dataset into Google Colab Pro, used the specific path for my custom object dataset and ran the command mentioned in your repo for running training but I notice that when the script begins to run it states the following: Train set size: 0 images, 0 identities.
I guess this is saying that I cannot read my dataset for some reason....do you know why this?
Any help would be appreciated. I've noticed that there's really know step by step guide anywhere online on how to create, annotate, classify, save and integrate a custom sequence to sequence dataset for deepsort object tracking.
Again, any help/guidance would be greatly appreciated.
John aka DataCowboy
The text was updated successfully, but these errors were encountered:
Hello Nwojke,
Thanks for all of your hard work.
I have created a custom dataset from scratch using sequence to sequence image frames from a video.
I have spent countless hours manually annotating all of the frames one by one, creating bounding boxes around each specified object and gave them their classification names.
I then saved the dataset as VOC file seeing that the images will be in jpg format and the annotation details will be in XML format.
Now here's where the issues occur, once I have clone your repo, I then upload the dataset into Google Colab Pro, used the specific path for my custom object dataset and ran the command mentioned in your repo for running training but I notice that when the script begins to run it states the following: Train set size: 0 images, 0 identities.
I guess this is saying that I cannot read my dataset for some reason....do you know why this?
Any help would be appreciated. I've noticed that there's really know step by step guide anywhere online on how to create, annotate, classify, save and integrate a custom sequence to sequence dataset for deepsort object tracking.
Again, any help/guidance would be greatly appreciated.
The text was updated successfully, but these errors were encountered: