By Jay Kakade 30 Dec, 2024
Collected at: https://www.techexplorist.com/researchers-develop-offline-speech-recognition-algorithm/95163/
As per a report, one out of four people are currently integrating with speech recognition. Since digital assistants like Siri, Alexa, and Google have secured a commonplace, their future roles are expected to become more prominent. Some predictions say that the devices connected to speech recognition will exceed the planet’s population.
Though the technology is evolving, speech recognition relies on an internet connection device. This limitation is due to the requirement of significant amounts of temporary random access memory (RAM), which large data centers often provide.
Researchers at the University of Copenhagen’s Department of Computer Science have developed a new algorithm to eliminate the need for an internet connection when using speech recognition on small devices. This algorithm allows devices to decode speech without substantial memory or internet access.
“Speech recognition fundamentally works by matching the small speech sounds we use to form words and sentences—known as phonemes—with a library of corresponding sounds,” explains Panagiotis Karras.
“Probabilities are calculated for matches and the subsequent combinations that go on to form our words and sentences. The most likely sequences are calculated and the software translates these sounds into text.“
Current algorithms demand increased memory the longer one speaks. Meanwhile, the new code employs a clever strategy: it “forgets” what it doesn’t need in real-time.
Co-developer and co-author Nassos Katsamanis says, “The algorithm, conceived by Panos and developed further by our team, does something entirely new. Unlike the existing gold standard algorithm used since speech recognition’s early days, our algorithm only stores a fraction of the processing data, serving as a set of ‘coordinates.’ With these, an entire sequence can be reconstructed, which makes speech recognition possible with significantly less RAM.“
Even though the algorithm sounds simple, its entirely new and unique code has enabled researchers to look for a patent. This algorithm reduces the need for memory and takes almost the same time to deliver recognition quality. However, with the increasing length and complexity of sentences, the demand for RAM increases.
“Certain small devices can already recognize and act based upon a few words without internet connectivity. For example, a smart home system can recognize keywords such as “turn on” or “turn off.” This is known as small-vocabulary speech recognition. With our algorithm, it will be possible to recognize more extensive instructions or, in principle, entire languages – without an internet connection. This is referred to as large-vocabulary speech recognition.“
Researchers assert that this innovation will open up new avenues in its applications, ranging from practical, security-related, and societal benefits along with its energy-saving potential. Since the code requires no internet connection, it eliminates the risks of data loss from an internet connection hack.
“This algorithm can help democratize language technology by making information more accessible. Making translation tools and speech assistants available regardless of internet access will allow more people to engage in society. In particular, it will help people without written language skills or those with physical disabilities, by enabling them to understand and influence societal decisions,” says Nassos Katsamanis.
Journal Reference
- Ciaperoni, M., Katsamanis, A., Gionis, A., Karras, P. (2024) Beam-search SIEVE for low-memory speech recognition. Proc. Interspeech 2024, 272-276, DOI: 10.21437/Interspeech.2024-2457
Leave a Reply