Artificial intelligence shoe cabinet using deep learning for smart home

Abstract

The recent development of smart home technology is making people’s residential life more affluent. Numerous advanced technologies are being applied to various home appliances and furniture including smart boiler, smart refrigerator and smart bed changing people’s everyday life gradually. The shoe cabinet is the furniture people pass through first when entering a house, and it was thought that if IoT function is attached to a shoe cabinet it will give much convenience to people as a component of the smart home just like smart boiler and smart refrigerator. On such a thought and trend, a small processor like Raspberry Pi was attached to the shoe cabinet at home to see the list of shoes, store shoes automatically and recommend right shoes for occasions. The shoes that the user wants to put into the shoe cabinet are stored automatically in the empty space of the shoe cabinet using x-y floater, and the shoe cabinet stores shoes by dividing the kinds and colors of shoes. And it is possible to see the status of the shoes in the shoe cabinet remotely using a mobile application. Moreover, the most appropriate shoes will be recommended when information on type of clothes worn and the destination are put in. The automatic storage of shoes was realized by controlling the input sensor and x-y floater with the Raspberry Pi attached to the shoe cabinet. Classification of shoe images was realized by putting in already classified shoe image data called ‘UT Zappos50K’ in the Deep Learning model made with Keras framework. Shoe recommendation service is made possible by entering the info on condition of user’s clothes and destination and recommending the shoes which score the highest among the shoes in the shoe cabinet using the prepared score chart.