Classification Modeling Method for Near-Infrared Spectroscopy of Tobacco Based on Multimodal Convolution Neural Networks

Figure 6

Model classification accuracy of different convolution kernel sizes.