Figure 6
From: Deep learning-enabled mobile application for efficient and robust herb image recognition
The confusion matrix of different networks on the testing set. (a) Small DNN with the data augmentation and network pre-train; (b) Small DNN without the data augment; (c) Small DNN without the network pre-train; (d) Small DNN using the network transfer; (e) Small DNN without the pre-train but using the network transfer.
