I’ve seen many AI projects fail quietly because of aggressive augmentation.
In our experiments, some augmentations helped. Many hurt. Blur, noise, and artificial effects reduced reliability.
Here’s the rule I’ve learned the hard way: if the augmented data doesn’t look like reality, the model won’t behave in reality.