Figure 6.

Performance evaluation (DepthConv-LSTM) with and without sparse learning: (a) without introducing new classes and (b) with the introduction of a smaller number of classes than existing.