File size: 1,013 Bytes
5942c3e
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
cff-version: 1.2.0
message: "If you use this work, please cite it using the metadata below."
type: software
title: "Multilingual Audio Intelligence System: Speaker Diarization, Speech Recognition and Translation"
authors:
  - given-names: "Prathamesh Sarjerao"
    family-names: "Vaidya"
    orcid: "https://orcid.org/0009-0000-8091-4256"
repository-code: "https://github.com/Prathameshv07/Multilingual-Audio-Intelligence-System"
license: CC-BY-NC-4.0
date-released: "2025-09-06"
abstract: "An end-to-end audio analysis pipeline that combines pyannote.audio speaker diarization, OpenAI Whisper multilingual speech recognition and a three-tier neural machine translation fallback (Opus-MT, Google Translate wrappers, mBART50/M2M-100), with a focus on Indian languages, served through a FastAPI web interface with waveform visualisation and cached demo mode."
keywords:
  - "speaker diarization"
  - "automatic speech recognition"
  - "machine translation"
  - "Indian languages"
  - "Whisper"
  - "pyannote.audio"