Results for "flat region"

Restricting updates to safe regions.

A wide basin often correlated with better generalization.

Flat high-dimensional regions slowing training.

Pixel-level separation of individual object instances.

The shape of the loss function over parameter space.

A narrow minimum often associated with poorer generalization.

Networks using convolution operations with weight sharing and locality, effective for images and signals.

Identifying and localizing objects in images, often with confidence scores and bounding rectangles.