Soo-Jong Lee

A weighted distribution of residual failure data using Weibull model

In analyzing the distribution of residual failure data using the Weibull distribution from the first detection of a failure to the completion of repairing failures, we have proposed the method of applying a weight to each failure. To put it concretely, we have suggested the method of applying a weight according to the importance of each failure within the analytic area in order to manage and measure efficiently the failure data found in the course of confirmation, verification, debugging, adding, and repairing functions collectively. The procedure is carried out after constituting a model switching system and installing a synthesis software package on the HANbit ACE ATM switching system under the development phase, which is the main node in building the B-ISDN for multimedia services.

An extraction method of lip movement images from successive image frames in the speech activity extraction process

In this paper, we propose an extraction method of lip movement images from successive image frames and present the possibility to utilize lip movement images in the speech activity extraction process of speech recognition phase. The image frames are acquired from the PC image camera with the assumption that facial movement is limited during talking. First of all, one new lip movement image frame is generated with comparing two successive image frames each other. Second, the fine image noises are removed. Each fitness rate is calculated by comparing the lip feature data as objectly separated images. It is analyzed whether or not there is the lip movement image through verification to the objects and three images which have higher rates in their fitnesses. As a result of linking the speech & image processing system, the interworking rate shows 99.3% even in the various illumination environments. It was visually confirmed that lip movement images are tracked and can be utilized in speech activity extraction process.

Speech Activity Detection with Lip Movement Image Signals

This paper describes an attempt to correlate lip movement visual information acquired via a camera with speech audio information acquired via a microphone from a human speaker in order to prevent audio created by external noise from being misrecognized as speech emitted by said speaker. Images of the face of a human speaker are acquired via a PC camera and are then separated into images that indicate lip movement and images that do not indicate lip movement. The data of lip movement image signals is saved in shared memory and shared with the speech recognition process. This data is analyzed by the speech activity detection process, which is a pre-processing step of sound recognition. We combined a speech recognition processor and an image recognizer, and the interworking function successfully operated at the rate of 99.3%.

Failure modeling for the ATM switching system under the development phase using Weibull distribution

This modeling analyzes the fluctuations for the whole failure data of past, present, and future ATM switching systems under the development phase, which is a core node in building the B-ISDN.