Language Identification

Our technology will help you to distinguish the language spoken. We will allow you to automatically route the valuable call to someone in your company who speaks the language or to a software that can analyze it, and we will make the process fast.

If you are in the security/defense sector, our language identification system will allow you to dig the most of telephone calls. 12 languages are already trained, you can train new language. Language Identification is useful for security/defence sector or call-center.

Here are some key features of "Language Identification":

· Input: Directory (waiting for incoming files) or list of files · Input formats: Linear coding (16 bits/8 bits), A-law, Mu-law · Sampling frequency: 8000 Hz · Channels: mono/stereo · Preprocessing: Filter that can cut-off technical signals (tones, fax, modem) and long silence parts. The length of input signal can be length limited to speed-up the processing. · Output: Speech file moved to a directory according language + log file, or one score file · Possibility to specify content of the log file · Possibility to store also cut out speech file · Languages: Readily available in the system: Arabic, English, Farsi, French, German, Hindy, Japanese, Korean, Mandarin, Spanish, Tamil, Vietnamese · Possibility to train a new language by user · Processing speed: 1/3 real-time

Note: In order to run the application you will need to get a temporary license key.