Middleware and Metrology for the Pervasive Future

The National Institute of Standards and Technology (NIST) has developed data and metrology tools for researchers, including common middleware for distributed sensor data acquisition and processing. The institute has provided researchers data sets to train and test classifiers that recognize words, speakers, events, and objects. It has provided metrics such as dynamic alignment for word error rates, speaker identification, and biometrics. It has also developed sensor-net middleware for data collection and synchronization of research efforts. Researchers at the institute have designed NDFS for localized high-bandwidth and multisensor data streams that consider large number of microphones and multiple video channels acquired on networked sensors and PCs. Researchers also made efforts at developing multiple-language bindings for NDFS, as it has been designed as an open middleware layer for distributed sensors and computations.