Speech-to-Text system

A key thing for voice interaction systems.
Now, there are many online services with similar functionality, but most of them lead to security issues (data leaks) and may cause delays in interaction (slow response).
Therefore, there is often a need to use something local.
This solution was written just for such cases.

Written in C, with libs: mpg123, alphacephei framework.
Capable to work on the regular servers, produces fast responses that suitable to build realtime dialogue systems.

Price: 350$ / 350 USDT
For purchase questions, please visit contact page.
A trial period with installation on your servers is provided (preferred Ubuntu 22.04 x64).

Basic features:

• Neural based, full locally system

  Doesn't depend on any online services, all data is processed locally

• Multilingual support

  There are open models for various languages

• Fine-tuning models

  There are tools for it

• Context dictionary

  Allows to defined (in request) a dictionary of accessible words

• Speaker identify

  Allows to generate vector for speaker identification

• Capable to work on regular servers

  for example: IBM x3550-M3

• Supported in FreeSWITCH

  Available in dialplan and scripts
  There's a module for integration (mod_sivr_asr)

• Preload and cache models

  Allows to save memory and improve performance

• Simple web api

  Easy integration with various applications

• Supports formats

  - wav
  - mp3
  - l16

• Supports os

  - Linux

--- Examples ---

Example #1 (simple request)

Request:
curl http://127.0.0.1:8801/v1/transcriptions -X POST -H "Authorization: Bearer secret" -H "Content-Type: multipart/form-data" -F language="en" -F smodel="small" -F file="@test.mp3"

Response (json):
{
 "text" : "hello world"
 }
        

Example #2 (with speakes identify)

Request:
curl http://127.0.0.1:8801/v1/transcriptions -X POST -H "Authorization: Bearer secret" -H "Content-Type: multipart/form-data" -F language="en" -F smodel="small" -F vmodel="default" -F file="@test.mp3"

Response (json):
{
 "spk" : [-0.644623, 1.023342, 2.575434, 0.623447, -0.602342, 1.0234234 -1.4824234 -0.021242, 0.824297, -0.152424, ... ],
 "spk_frames" : 81,
 "text" : "hello world"
}