IMTC is working
with researchers in the Center for Signal and Image Processing at
Georgia Tech to implement novel algorithms to reduce background
noise in spoken audio. Our approach provides results superior to
standard approaches and employs a more intuitive user interface.
The system is designed as a DirectX plug-in to allow it to function
in a variety of Windows based applications. Additional applications
in such fields as wireless telephony are being explored as well.
Preliminary experimental results were first presented at the 142nd
Meeting of the Acoustical Society of America on December 6, 2001. Further
results detailing improvements in speech recognition and word spotting
were presented at ICME 2003 (International Conference on Multimedia & Expo)
and are published in the conference proceedings. (Download
the ICME paper in PDF format © ICME 2003)
Speaky Clean – Automatic Noise Remover v 0.95 beta
Speaky Clean – Automatic Noise Remover packages IMTC’s noise reduction technology into a simple software tool that can run as a stand-alone application for Windows and as a DirectX audio plug-in compatible with popular audio and video editing applications.
Speaky Clean is currently open for Beta testing. Please share your experiences with our development team!
Email: speakyclean@imtc.gatech.edu
Download Speaky Clean v. 0.96 beta (.exe – 3.65 MB)
(updated 10/26/2006)
click on the thumbnail below for a larger view (JPG: 640 x 480)
|
Image of the Speaky Clean interface
Examples
- Example #1 : Air Conditioner
- unprocessed audio file - wav
format (file download - 216k)
- processed noise reduced file - wav
format (file download - 216k)
- Example #2 : Cafeteria
- unprocessed audio file - wav
format (file download - 271k)
- processed noise reduced file - wav
format (file download - 271k)
- Example #3 : Car
- unprocessed audio file - wav format (file download - 797k)
- processed noise reduced file - wav
format (file download - 797k)
- Example #4 : Clatter
- unprocessed audio file - wav
format (file download - 132k)
- processed noise reduced file - wav
format (file download - 132k)
- Example #5 : Fan
- unprocessed audio file - wav format (file download - 160k)
- processed noise reduced file - wav
format (file download - 160k)
- Example #6 : Interview
- unprocessed audio file - wav
format (file download - 753k)
- processed noise reduced file - wav
format (file download - 753k)
- Example #7 : TIDIGITS (white)
- unprocessed audio file - wav
format (file download - 46k)
- processed noise reduced file - wav
format (file download - 46k)
- Example #8 : TIMIT
- unprocessed audio file - wav
format (file download - 364k)
- processed noise reduced file - wav
format (file download - 364k)
|
]
[
