HEAD acoustics continues to excellent measurement and evaluation options within the subject of speech and communication high quality, increasing from an unlimited vary of {hardware} platforms and instruments – together with its HMS II.3 fullband synthetic head – all the way in which to devoted software program that adjusts to the quickly-evolving wants on this subject. The German firm continues to increase VoCAS, a complicated software program instrument for efficiency analysis of computerized speech recognizers, addressing actual life situations within the lab.
The advantage of good digital assistants and their acceptance by the customers rely on how nicely they acknowledge spoken instructions and reply to them in real-life environments – and the idea for good efficiency are computerized speech recognizers. VoCAS (Voice Management Evaluation System) from HEAD acoustics reveals speech recognizer efficiency with unprecedented ease and accuracy, and is used to benchmark completely different speech recognition techniques.
The last word instance of the German firm’s skill to deal with goal sign analysis in advanced eventualities, the VoCAS software program objectively exams and evaluates speech recognition efficiency in good dwelling purposes, autos, or smartphones – to call just some – emulating a sensible surroundings for every respective software. It’s a part of HEAD acoustics’ refined measuring and playback framework that mimics real-life conditions as carefully as potential, incorporating vital elements in measuring speech recognizer high quality with a novel, complete method. VoCAS considers the acoustic surroundings, background noise, completely different talkers, completely different languages, and even accents – all elements that remained till know uncorrelated.
As a result of consideration to element means consideration to variation, this true-to-life diligence additionally entails utilizing HEAD acoustics’ HMS synthetic heads, outfitted with a synthetic mouth utilizing specifically developed loudspeakers simulating the human mouth sound subject. And VoCAS permits the exact particular person calibration of every synthetic mouth. This method permits mimicking the pure voice transmission path, which is the one means to make sure a sensible human voice sound subject technology.
With VoCAS’ customizable and reproducible take a look at sequences tailored to the respective software, customers acquire important data to optimize and validate the speech recognition efficiency, with as much as 4 simultaneous talker positions – even in noisy, reverberant environments.
The HMS synthetic heads are built-in with HEAD acoustics {hardware} platform labCORE, permitting VoCAS to run advanced take a look at eventualities with as much as 4 talkers concurrently, automated and calibrated speech recordings, and a simulation of various room environments and noise situations. This makes VoCAS ideally suited to benchmarking numerous acoustical system interfaces together with speech recognition techniques. In fact, VoCAS additionally permits customers to import present speech materials and lower, filter, and alter audio recordsdata to outlined speech ranges.
With the continual function enlargement and updates, VoCAS additionally makes it simple for customers to understand the total potential of the software program and its ecosystem. For instance, its in depth take a look at automation now contains user-programmable take a look at sequences, totally computerized equalization, automated analysis, and the statistical evaluation of detection and error charges. VoCAS additionally combines a database software that integrates right into a complete setup for the synchronized and calibrated playback of speech and spatially recorded knowledge.
Launched in 2016 by HEAD acoustics, and since then repeatedly improved, VoCAS is now in a position to take a look at and consider any speech recognition system systematically and below sensible, reproducible situations. The truth that it allows automated, advanced, and sensible take a look at eventualities with as much as 4 talkers, background noise, and reverberation and echo, make it ideally suited to optimizing and benchmarking units using or linked to speech recognition techniques.
HEAD acoustics GmbH is without doubt one of the world’s main firms providing holistic options for sound and vibration evaluation, having fun with world recognition within the telecom market resulting from its pioneering position within the growth of {hardware} and software program for the measurement, evaluation and optimization of voice and audio high quality in addition to customer-specific options and providers.
HEAD acoustics’ vary of providers covers sound engineering for technical merchandise, investigation of environmental noise, speech high quality engineering in addition to consulting, coaching and assist. The corporate is headquartered in Herzogenrath, close to Aachen, in Germany, and has subsidiaries in China, France, Italy, Japan, South Korea, the UK and the USA in addition to quite a few gross sales companions worldwide.
www.head-acoustics.com
