We are surrounded by objects with personal memories. However, it is hard for people with visually impairments to keep remembering and recognize such connections without vision.
POR (Personal Objects Recognizer) aims to help them regain those connections between objects and memories.
The information of “T-shirt” and “mug” are not enough. We want to know which t-shirt and mug.
Users are able to add the information by themselves, and it allows them to get the information what they want to know.
Users themselves train the POR by collecting data and register their personal objects with their comments. It is based on deep learning technology.
POR guides blind users to take good photos of a targeted object by utilizing AR toolkit of iOS to create 3D point cloud around the targeted object.