Techware: Speech recognition software and resources on the web [Best of the Web]

In this issue, "Best of the Web" presents online resources available to tackle the problem of speech recognition. Automatic speech recognition turns spoken audio into a sequence of words. It is an extraordinarily broad and multidisciplinary field, drawing primarily from statistical signal processing, machine learning, and linguistics. Although speech-recognition-oriented applications are essential to the success of the technology and its acceptance in the marketplace, we have excluded them from the scope of this article. We will focus instead on resources that are useful for a beginning apprentice to build the technology that will turn sounds into words.