Segmentation labeling makes more sense for us, Because we can train the model on shapes.
Shape of an ear being close to the shape of the trunk is highly possible. Or other combinations of shapes being together should give us a good handle on if it’s an elephant or not.
Need your thoughts to fix the labeling mechanism we’ll be using
@suryamurugan @johnkaramchand @Thanmay47
Labeling and Use cases.
Segmentation labeling makes more sense for us, Because we can train the model on shapes. Shape of an ear being close to the shape of the trunk is highly possible. Or other combinations of shapes being together should give us a good handle on if it’s an elephant or not.
Need your thoughts to fix the labeling mechanism we’ll be using @suryamurugan @johnkaramchand @Thanmay47