Singapore "I" Vessel Assessment (SIVA)
Computer-aided Integrated Platform for Large-scale Non-invasive Observation of Cardiovascular Disorders Using Retina Image Analysis
In collaboration with SERI, National University of Singapore (NUS) computer scientists developed the Singapore “I” Vessel Assessment (SIVA) software to extract retinal vascular structure and derive quantitative measures from retinal images to describe the retinal vessels' characteristics.
The SIVA software is a user-friendly system with accurate and robust algorithms that can measure the vascular or blood-vessel structure in retinal images automatically.
SIVA is also flexible and intuitive in gathering feedback to enhance the accuracy of vessel measurement and description. SIVA can automatically compute a spectrum of retinal vascular parameters including retinal vascular caliber, tortuosity, branching angle, fractal dimension and junction exponent deviation from retinal fundus photographs to quantify the retinal vasculature. Other automation of SIVA includes retinal vasculature tracing, vessel type classification (venule or arteriole), optic disc detection and position the measured grid following the Atherosclerosis Risk in Communities (ARIC) Study protocol.
Robust Speaker Recognition RSR2015 Speech Database
Speech corpus for text-dependent voice biometrics software evaluation
The RSR2015 speech database is a collection of voice data from 300 speakers (143 female and 157 male), with 657 utterances from each speaker. The database contains 151 hours of speech recording in total and occupies approximately 50 GB of storage space. The database has been created for use as a corpus to evaluate the performance of speaker verification engines.