Is this model is applicable for transfer learning and continuous Learning?

[himabinduyeddala]

Hi@ihciah can you please tell me your deep fashion model is applicable for continuous learning and transfer learning for inclusion of new categories.

Best, Himabindu Y

[ihciah]

Maybe yes. Add categories to the last layer doesn't hurt the feature layer, which is the second last layer. The feature extractor should work fine when you add new categories. The last fully connecting layer should be re-initialized and trained. The other paramters can be loaded from the old model.