https://github.com/bilalhameed248/whisper-fine-tuning-for-pronunciation-learning

Fine Tuning of Whisper Speech To Text Base Model For Pronunciation Learning
https://github.com/bilalhameed248/whisper-fine-tuning-for-pronunciation-learning

deep-learning deep-neural-networks dnn fine-tuning openai pronunciation python seq2seq speech speech-recognition speech-synthesis speech-to-text whisper whisper-ai

Last synced: about 2 hours ago
JSON representation

Fine Tuning of Whisper Speech To Text Base Model For Pronunciation Learning

• Host: GitHub
• URL: https://github.com/bilalhameed248/whisper-fine-tuning-for-pronunciation-learning
• Owner: bilalhameed248
• Created: 2023-07-16T14:14:54.000Z (over 1 year ago)
• Default Branch: main
• Last Pushed: 2024-04-23T15:12:51.000Z (7 months ago)
• Last Synced: 2024-06-08T20:01:51.652Z (5 months ago)
• Topics: deep-learning, deep-neural-networks, dnn, fine-tuning, openai, pronunciation, python, seq2seq, speech, speech-recognition, speech-synthesis, speech-to-text, whisper, whisper-ai
• Language: Jupyter Notebook
• Homepage:
• Size: 207 KB
• Stars: 0
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

        
Whisper Fine-tuning for Pronunciation Learning




In this project, I undertook the task of fine-tuning a whisper speech to text base model to enhance pronunciation learning, particularly focusing on broken words or fragmented speech segments. The primary objective was to develop a robust system capable of accurately transcribing whispered speech, especially in scenarios where words are partially uttered or fragmented. Leveraged advanced transfer learning techniques and deep learning architectures to achieve an impressive accuracy rate of nearly 95%. Collaborated with educators to integrate the model into language learning applications, demonstrating a commitment to leveraging technology for educational enhancement.