The use of bioelectric signals has the disadvantage that users have to wear the electrodes on their body. This can be intrusive, uncomfortable, and not aesthetically pleasing. Such a device also suffers from being dependent on the ambient conditions such as temperature and humidity, and is sensitive to body sweat and movement artifacts, thus limiting the possible applications.

Recognition of commands from the video of the user has the advantage because recording the body movements do not require direct contact with the body and can generally be nonintrusive. This chapter discusses the technologies for recognizing hand gestures from videos and the challenges associated with the real-time implementations, and also describes an example of the implementation of the technology. The current research trend and future directions are also covered in this chapter.