Get on-line emotion information
We have developed a vibrotactile chair, FEAT TM, (Tactile Facial Expression Appearance System); a social interface for the visually impaired. With this chair the visually challenged person can get on-line emotion information from the person he/she is heading to. This greatly enhances communication ability and improves the quality of social life of the blind.
To help the visually challenged to understand, interpret and interact with normal people, a tactile video system has been built. The system transfers and interpret dynamic facial expressions into meaningful tactile sensory information.
The tactile video system consists of three subsystems:
The imaging system is used to capture facial expressions. Sensed signals may be obtained from visible light, infrared light, or a combination of them. This system could also be a video camera array specially designed to satisfy some requirements, e.g. stereo video.
The image understanding system contains two kinds of image understanding modules, Lowlevel visual processing (LLVP) and High-level visual processing (HLVP). The image(s) are processed through independent LLVP modules corresponding to different visual cues, in parallel. These cues representing color, motion, shape, range, texture, and other visual primitives are input to HLVP modules. It is the HLVP module that extracts high-level descriptions of images like facial expression, emotion etc. The tactile actuator systems, normally attached to the the back, are used to transform the extracted visual information into pressure signals
