Disentanglement of Color and Shape Representations for Continual Learning

Cited by: 0|Bibtex|Views5|Links

Abstract:

We hypothesize that disentangled feature representations suffer less from catastrophic forgetting. As a case study we perform explicit disentanglement of color and shape, by adjusting the network architecture. We tested classification accuracy and forgetting in a task-incremental setting with Oxford-102 Flowers dataset. We combine our m...More

Code:

Data:

Your rating :
0

 

Tags
Comments