Ddbn May 2026

Ddbn May 2026

A semi-supervised classifier that combines the generative power of Deep Belief Nets (DBN) with the discriminative power of backpropagation.

Primarily used for visual data classification and scene recognition (e.g., indoor environment classification).

Replaces standard Feature Pyramid Networks (FPN) with dual detection branches (e.g.,