Accessibility ultimately aims at an inclusive society for people with disabilities, senior citizens, people with literacy problems, and everyone who has access challenges in their daily lives, education, and employment. Throughout history, many now popular and widespread technologies were originally invented to help support people with disabilities, including keyboards, voice synthesis, optical character recognition, and many others. IBM Research – Tokyo played an important role in creating and applying such technologies as voice recognition and synthesis, digital Braille, voice-based Web access, and crowd-sourcing for accessibility enhancements. We will continue exploring new technologies to tackle the new challenges in our society. Recent work can be classified into three areas: Accessibility” for people with disabilities and seniors, Speech Technologies for advanced voice analysis, and Social Computing for inclusive enterprises and societies.
- Accessibility for Persons with Disabilities and Senior Citizens Accessibility research is creating new technologies to assist persons with disabilities and senior citizens to actively participate in society. IBM Research – Tokyo has been working in this area for many years. The digital Braille authoring and sharing system developed in the 1980s is still in use as the basis of the current de facto standard systems. IBM Home Page Reader in 1997 was the world’s first practical voice Web browser and contributed significantly towards later assistive technologies and Web accessibility standards. The accessibility evaluation tool aDesigner was open sourced in 2007 and became the basis of the Japanese Government official accessibility initiative in 2011. We are now working to explore new technologies for video accessibility, digital book accessibility, usability for seniors, and summarization technologies based on analytics technologies. One of our most promising new technologies is called Sasayuki, a system to provide context-and-task-based support.
- Speech Technologies Automatic speech recognition (ASR) is a technology which converts an utterance into text by analyzing human voice input with a computer. In 1998, IBM Research-Tokyo commercialized IBM ViaVoice, a Japanese large vocabulary continuous speech recognition (LVCSR) software package, prior to competitors. Back then the LVCSR software had focused on analyzing well-pronounced speech uttered in a noiseless environment. After that, the technologies have been improved and our current target for analysis includes spontaneous speech like that which occurs in conversations between people. Research findings have been utilized in various business areas such as call center monitoring, smart phones, and car navigation systems. IBM Research-Tokyo continues to research various speech technologies focusing on speech recognition, synthesis, and analytics in a practical environment.
- Social Computing Social Computing is a cross-disciplinary area between computer science and social science, and we are focusing on applications of crowd-sourcing technologies. Crowd accessibility is a technological concept to apply crowd sourcing for accessibility improvements. Crowd Card is a new project to apply gamification strategies to increase participation in social applications and also help enterprises fully utilize the employees’ eclectic skills. Senior Cloud is a project supported by JST (an agency of the Japanese government) to explore new technologies to discover and use the skills of senior citizens. We are also collaborating with other research groups such as social analytics teams.