Automatic Speech Recognition

Speech-to-text model trained on custom datasets.

PyTorchAudioMachine Learning

Problem Statement

Converting raw audio acoustic signals into precise text transcripts under variable audio conditions.

Role & Responsibilities

Role: Machine Learning Developer

Technical Solution

Trained and evaluated end-to-end acoustic speech-to-text models using custom dataset preprocessing and PyTorch pipeline.

Technologies & Tools:

  • PyTorch
  • Audio
  • Machine Learning
  • Python

Results & Impact

Demonstrated low word error rate (WER) performance on evaluation datasets.

View Repository on GitHub

Explore Other Projects →