Data-Augmentation Analysis

Home

Discover

Bookmarks

Keep Up With Latest Trending Papers. Computer Science, AI and Machine Learning and more.Subscribe

A Data-Augmentation Is Worth A Thousand Samples: Exact Quantification From Analytical Augmented Sample Moments

Data-augmentation (da) is known to improve performance across tasks and datasets.

We derive several quantities in close-form, such as the expectation and variance of an image, loss, and model output under a given da distribution.

Those derivations open new avenues to quantify the benefits and limitations of da.

For example, we show that common common data-augmentation require tens of thousands of samples for the loss at hand to be correctlyestimated and for the model training to converge.

We show that for a training loss to be stable under da sampling, the model s saliency map (gradient of the loss with respect to the model s input) must align with the smallest eigenvector of the sample variance under the considered da augmentation, hinting at a possible explanation on why models tend to shift their focus from edges to textures.

Authors

Randall Balestriero, Ishan Misra, Yann LeCun

◐ Latest Activity

◐ Recommended Follows