Signing you in…

Data augmentation: policies, mixup and regularization

Augmentation: grow the dataset without new images

A model sees a cat upright and may fail on an upside-down cat—it lacks invariance. Augmentation fixes this at training time: each epoch shows the same image under different transforms. A dataset of 1,000 images with rich augmentation behaves like a nearly endless stream of diverse examples.

Content is available with subscription.
Get full access to all courses on the platform for one year with a single payment.
Unlike other platforms that charge per course, here you get everything for one price, and after one year of use there will be no automatic charge for the following year.