Multisensory Culinary Image Classification based on SqueezeNet and Support Vector Machine

Food tourism can be a competitive advantage for Indonesian tourism. With diverse cultures and ethnicities, Indonesia certainly has local culinary characteristics not found in other regions. Although culinary tourism research has been conducted, not many have examined culinary tourism from multisensory experience analysis through a big data approach to obtain datasets that can be processed to visualize tourist experiences. This study aims to classify the multisensory experienced by tourists when doing culinary tours utilizing deep learning SqueezeNet for image extraction and Support Vector Machine with Linear Kernel for image classification. The primary dataset consisted of three image classes: Vibes, Place, and Food. This study succeeded in classifying multisensory culinary images with an accuracy rate of 98.6%.