Assistive technologies, such as devices to help perform tasks, cognitive aids, mobility aids, physical modifications in the built environment and closed captioning, help to improve or maintain a person’s ability to complete day-to-day tasks. For example, assistive technologies can be helpful for people with disabilities or the elderly, enabling them to work around challenges they may encounter. The development of assistive technologies depends on datasets, which are used for training, testing or validation. Furthermore, making such datasets widely available can help advance the field of assistive technologies. One area in which assistive technologies can be useful is in helping people who are visually impaired. For example, electronic mobility aids use ultrasonic waves that reflect off objects in front of people, letting them know what is ahead. At the Division of Industrial Art, Faculty of System Design, Tokyo Metropolitan University, in Japan, Associate Professor Tetsuaki Baba is using a deep learning approach to develop assistive technology for visually impaired people. As a foundation for this work, he and his team have developed a dataset for developers to create deep learning vision-based applications for visually impaired people.