The Engineering Company for the Development of Digital Systems

Contact Info
12A Haroun,Doqi, Giza Governorate, Egypt
info@rdi-eg.com
+20 2 37 49 94 63 +20 2 37 49 55 66 +20 2 37 49 95 61

Speech to text Arabic

Converting your audio data into valuable text.
Speech to text Arabic - Convert video to text تحويل الصوت الى نص - تحويل الفيديو الى نص

Kateb is an Automatic Arabic Speech to Text (STT) software solution tailored for Arabic voice recognition. Kateb transcribes recorded Arabic audio/video files into fully editable and searchable text files. Besides supporting Modern Standard Arabic (MSA), Kateb supports the Egyptian Dialect; as well as the Saudi Dialect. So Kateb’s innovative technology guarantees its users with one of the best Speech-to-Text experiences, delivering high, fast, and accurate recognition.

Main Features

Accurate and Reliable

RDI uses the most advanced deep-learning neural network algorithms for the speech-to-text engine. Also, RDI uses multiple recognition layers to guarantee the most accurate  transcription.

Supports Multiple Sampling Rates

Kateb supports different audio sampling rates. 16KHZ (16 bit per sample) sampling rate is recommended; and can easily adapt its engines to fit low sampling rate audio files based on customers’ needs.

Fast Transcription

Automatic speech to text at high-speed recognition that transcribes voice at least 5 times faster than regular typing speed. As well as, faster recognition rates are available using hardware with higher specifications; or at using custom models that can be built based on customer’s needs.

Supports Multiple Input Formats

Kateb supports major audio/video file formats as well as all types of files that follow the FFmpeg formats. This makes it easy for developers to pass different files in different formats directly without the need to convert them into specific formats(s).

Supports Multiple Output Formats

Transcribed text can be saved in various formats, providing you with more than just text. The output transcription can include the timing of the words, silence, music, and entire segments. As well as the recognition certainty level for your content.

Easy Editing & Reviewing

The online service provides instant synchronization between speech spans and resultant text. End users can easily modify any transcription mistakes.

Comprehensive Generic Vocabulary

Certainly Kateb can recognize a wide range of unusual Arabic words thanks to its huge vocabulary library. RDI updates Kateb’s library periodically, to ensure constant development and continuous addition of the new terminology.

Diarization Modules

In addition to the voice recognition engine, Kateb detects speech, silence, and music in input source files through detection modules, whereas segments of speech can be specified with their accurate time spans.

Various Linguistic Input

Kateb supports Modern Standard Arabic (MSA), as well as Egyptian and Saudi Dialects. Clearly, the engine can be adapted easily to new dialects based on customers’ needs.

Why Kateb?

Domain & Environment Adaptation

Kateb can be adapted easily at the domain and environment levels. RDI can create custom models tailored for your needs, guaranteeing high recognition accuracy.

Tailored for Arabic

Kateb contains a set of models specially designed to serve the Arabic language, making the recognizer outperforms its general-model counterparts.

Productivity Improvement

Kateb improves your personal and business productivity as it is 5 times faster than typing with the possibility of faster recognition rates using higher hardware specifications.

Your Privacy Matters

Kateb respects customer privacy and does not share the users’ input/output files with third parties. We can set up the solution environment at your side, train your employees, and provide a full customer support service.