
Announcements
HI @admin052945
AI Builder is using Azure Custom Vision in the backend for training Object detection model. Regarding to image limitation, you can refer to the doc https://docs.microsoft.com/en-us/azure/cognitive-services/custom-vision-service/limits-and-quotas . We're using the S0 (standard) tier, so the max training images per project is 100k and max Tags/Labels (objects) are 500 per project.
If you have a large amount of objects you could try to break them down into different models. Also when we say min 15 images for each object, and you have 3 objects, it doesn't mean each of your image should only have 1 object, you can still prepare in total 15 images, each contains 3 objects to train a model. We recommend 50+ per object though.
Let me know if you have more questions.
Thanks
Norah