Posts tagged speech recognition
Articles and news from the latest research reports.
Researchers demonstrate information processing using a light-based chip inspired by our brain
In a recent paper in Nature Communications, researchers from Ghent University report on a novel paradigm to do optical information processing on a chip, using techniques inspired by the way our brain works.
Neural networks have been employed in the past to solve pattern recognition problems like speech recognition or image recognition, but so far, these bio-inspired techniques have been implemented mostly in software on a traditional computer. What UGent researchers have done is implemented a small (16 nodes) neural network directly in hardware, using a silicon photonics chip. Such a chip is fabricated using the same technology as traditional computer chips, but uses light rather than electricity as the information carrier. This approach has many benefits including the potential for extremely high speeds and low power consumption.
The UGent researchers have experimentally shown that the same chip can be used for a large variety of tasks, like arbitrary calculations with memory on a bit stream or header recognition (an operation relevant in telecom networks: the header is an address indicating where the data needs to be sent). Additionally, simulations have shown that the same chip can perform a limited form of speech recognition, by recognising individual spoken digits (“one”, “two”, …).
Google simulates brain networks to recognize speech and images
This summer Google set a new landmark in the field of artificial intelligence with software that learned how to recognize cats, people, and other things simply by watching YouTube videos (see “Self-Taught Software“).
That technology, modeled on how brain cells operate, is now being put to work making Google’s products smarter, with speech recognition being the first service to benefit, Technology Review reports.
Google’s learning software is based on simulating groups of connected brain cells that communicate and influence one another. When such a neural network, as it’s called, is exposed to data, the relationships between different neurons can change. That causes the network to develop the ability to react in certain ways to incoming data of a particular kind — and the network is said to have learned something.
Researchers at the Norwegian University of Science and Technology (NTNU) are combining two of the best-known approaches to automatic speech recognition to build a better and language-independent speech-to-text algorithm that can recognize the language being spoken in under a minute, transcribe languages on the brink of extinction, and make the dream of ever present voice-controlled electronics just a little bit closer.
Achieving accurate, real-time speech recognition is no easy feat. Even assuming that the sound acquired by a device can be completely stripped of background noise (which isn’t always the case), there is hardly a one-to-one correspondence between the waveform detected by a microphone and the phoneme being spoken. Different people speak the same language with different nuances – accents, lisps and other articulation defects. Other factors such as age, gender, health and education also play a big role in altering the sound that reaches the microphone.
The NTNU researchers are now pioneering an approach that, if it can be fully exploited, may lead to a big leap in the performance of speech-to-text applications. They demonstrated that the mechanics of human speech are fundamentally the same across all people and across all languages, and they are now training a computer to analyze the pressure of sound waves captured by the microphone to determine which parts of the speech organs were used to produce a phoneme.
Infants’ Recognition of Speech More Sophisticated Than Previously Known
ScienceDaily (July 17, 2012) — The ability of infants to recognize speech is more sophisticated than previously known, researchers in New York University’s Department of Psychology have found. Their study, which appears in the journal Developmental Psychology, showed that infants, as early as nine months old, could make distinctions between speech and non-speech sounds in both humans and animals.
A new study shows that infants, as early as nine months old, could make distinctions between speech and non-speech sounds in both humans and animals. (Credit: © ChantalS / Fotolia)
"Our results show that infant speech perception is resilient and flexible," explained Athena Vouloumanos, an assistant professor at NYU and the study’s lead author. "This means that our recognition of speech is more refined at an earlier age than we’d thought."
It is well-known that adults’ speech perception is fine-tuned — they can detect speech among a range of ambiguous sounds. But much less is known about the capability of infants to make similar assessments. Understanding when these abilities become instilled would shed new light on how early in life we develop the ability to recognize speech.
In order to gauge the aptitude to perceive speech at any early age, the researchers examined the responses of infants, approximately nine months in age, to recorded human and parrot speech and non-speech sounds. Human (an adult female voice) and parrot speech sounds included the words “truck,” “treat,” “dinner,” and “two.” The adult non-speech sounds were whistles and a clearing of the throat while the parrot non-speech sounds were squawks and chirps. The recorded parrot speech sounds were those of Alex, an African Gray parrot that had the ability to talk and reason and whose behaviors were studied by psychology researcher Irene Pepperberg.
Since infants cannot verbally communicate their recognition of speech, the researchers employed a commonly used method to measure this process: looking longer at what they find either interesting or unusual. Under this method, looking longer at a visual paired with a sound may be interpreted as a reflection of recognition. In this study, sounds were paired with a series of visuals: a checkerboard-like image, adult female faces, and a cup.
The results showed that infants listened longer to human speech compared to human non-speech sounds regardless of the visual stimulus, revealing the ability recognize human speech independent of the context.
Their findings on non-human speech were more nuanced. When paired with human-face visuals or human artifacts like cups, the infants listened to parrot speech longer than they did non-speech, such that their preference for parrot speech was similar to their preference for human speech sounds. However, this did not occur in the presence of other visual stimuli. In other words, infants were able to distinguish animal speech from non-speech, but only in some contexts.
"Parrot speech is unlike human speech, so the results show infants have the ability to detect different types of speech, even if they need visual cues to assist in this process," explained Vouloumanos.
Source: Science Daily