PRODUCTS
|
 HOME
Welcome to GritTec laboratory specializing on research and development of the unique algorithms and technologies in the field of digital signal processing (DSP).
We develop and market quality of DSP software, algorithms and solutions
implemented as floating, fixed point C language models or optimized on a range of DSP platforms.
GritTec's research is focused on these key points:
- Speech enhancement & concealment;
- Voice biometric technologies;
- Speech recognition;
- Methods of adaptive filtration;
- And other speech & audio technologies.
|
|
OVERVIEW PRODUCTS
BIOMETRIC TECHNOLOGIES
Automatic text independent speaker identification system
is intended for automatic identification of a speech signal of unknown
voice by comparing with "speaker cards", existing in the
database of system. ...
Automatic voice transcription system
is used for making phonetic transcription of a speech signal of unknown
voice and its language identification. ...
SPEECH ENHANCEMENT
Single-microphone solution
of speech enhancement can be used for reduction of external hindrances and
broadband color noises in speech signal. Controlled level of noises
suppression in speech is up to 60 dB.
...
Multi-microphone solution
can be effectively used when noise from many different sources needs
to be canceled. In its simplest form the technique uses two microphone solution. Speaker is recorded synchronous with the source
of hindrance (noise). The first microphone bases near by speaker, but the
second one bases near by source of hindrance or noise. ...
Microphone Array solution
of speech enhancement can be used for reduction of external hindrances
and ambient noises in speech source signal. In its simplest form the
technique uses two array microphone configurations. By using
acoustical beam-forming techniques, the microphone array can
"focus" on the speaker's position. ...
AUDIO-SPEECH EFFECTS
Time stretching technology
is used for time scale modification (changing the speed or duration) of
an audio signal without affecting its pitch.
Principle of functioning algorithm is based on time-domain modification using SOLA technology, such as Waveform Synchronized Overlap-Add (WSOLA). ...
Sample Rate Converter
is technology used for changing sampling rate in speech and audio signals.
Principle of functioning algorithm is based on methods of interpolation.
...
SIGNAL PROCESSING
Voice Activity Detector (VAD)
is used for detection silence intervals. Principle of functioning of
detector is based on methods of adaptive filtrations, allowing tracing
dynamically appearance of fragments of noise signal. Depending on
conditions of the channel, mips and quality detection, it’s possible
different realizations of variants Voice Activity detectors. It allows
using them effectively on the background of broadband noises and
channel distortion. ...
Phone Talk Detector (PhTD)
is used for detection of speech intervals. Such detector can be used
in devices, where it is important to detect exactly fragments of
speech signal. For instance, in telephone systems, when it is
necessary to detect moment of the beginning of talk between speakers
on telephone line and its duration. ...
Double Talk Detector (DTD)
can be used for estimate degree of similarity between channels. Double
Talk Detector effectively can be used at the suppression of echo in
devices VOIP telephony (with echo cancellers). It's developed two
modification of Double Talk detectors. ...
Acoustic Clipping Detector (ACD)
is used for estimation degree of clipping signal, appearing because of
the excess of maximum amplitude level of signal. The
restrictions of the maximum level of signal amplitude can be added by
mike or channel itself or both of them simultaneously.
Acoustic Clipping Detector can be effectively used in devices IP
telephony, Call service centers, where it’s necessary constantly
monitoring of the incoming clipping signal (acoustic shocks). ...
|
|
|
|
