Automatic speech recognition (ASR) technology is an integral feature of many IVR systems. ASR enables a user to speak their answers as opposed to numerically entering them, permitting for much greater depth and quality of answers offered.
A study conducted by researchers from the Hammill Institute on Disabilities utilized ASR technology to track babies’ early language development by recording, transcribing and analyzing the amount and type of language queues they were exposed to within their home environments.
The findings were in keeping with previous studies, showing that children of the most talkative parents were the ones producing more vocalizations and conversational turns. This learning and speech pattern continued and, in fact, increased month over month.
The data reflects the idea that children’s exposure to language at home significantly affects the rate at which they learn to talk. Analyzing the data, researchers also found that there is no “normal” when it comes to home learning environments, and that each child’s situation varied in terms of their exposure to language.
Other interesting findings? Adult females provided the majority of the words heard by the child per day, but this could be due in part to the fact that the mothers included in the study reported spending significant time at home with their baby, more than most average mothers.
The technology utilized in this study was quite unique. The ASR was able to accurately distinguish between adult and child vocalizations. The fact that the speech recognition engine could easily and seamlessly separate adults and children’s voices offers a variety of possibilities for those collecting and processing data based on vocalizations.
Referring back to my write-up from last week, if speech recognition technology is able to separate adult and children’s voices, it could also feasibly measure less quantifiable aspects of the human voice that are equally important in this type of data collection, specifically emotion and intonation.
The bottom line is that ASR technology enables research advancement through its use. Research on “autism, effects of television, language learning in linguistically diverse environments and language intervention” are all possibilities that become realistic through the use of this type of system.