Skip to content

Latest commit

 

History

History
45 lines (38 loc) · 2.16 KB

README.md

File metadata and controls

45 lines (38 loc) · 2.16 KB
Raw

Dynamic Time Warping (DTW) Speech Recognition

A speech recognition system that uses feature extraction and dynamic time warping (DTW) to identify words and to find the most similar speaker.

Features

  • Word Recognition: Identifies words spoken by different speakers using a weighted voting mechanism across multiple feature types.
  • Speaker Similarity Detection: Determines the most similar speaker to a target speaker based on feature comparisons.
  • Dynamic Time Warping (DTW): Measures similarity between time-series data with a focus on minimizing cost paths.
  • Multiple Feature Types: Supports MFCC, LPC, and Mel Spectrogram features for diverse and accurate audio analysis.
  • Confidence and Threshold Handling: Ensures reliable predictions by applying confidence thresholds and cost checks.

Setup

  1. Clone the repository:
git clone https://github.com/mradovic38/dtw-speech-recognition
cd dtw-speech-recognition
  1. Install dependencies:
pip install -r requirements.txt

Usage

  1. Create a directory of audio files of different speakers saying the same group of words.
  2. Name your audio files as word-speaker.wav (ex. down-mark.wav).
  3. For details on how to run the project on your data, refer to the run.py file. It contains examples and explanations for word recognition and speaker similarity detection.

Customization

  • Adjust feature weights in word_recognition.py to prioritize specific features.
  • Modify confidence and cost thresholds to suit different datasets.

How it works

Feature Extraction:

Extracts audio features such as MFCC, LPC, and Mel spectrograms from input audio files.

Dynamic Time Warping (DTW):

Computes similarity between time-series data (input audio vs. database).

Word Recognition:

Aggregates feature-based predictions using weighted voting across all three feature types for reliable word identification.

Speaker Similarity:

Compares speakers by averaging DTW costs across shared vocabulary based on a passed feature extraction alghoritm.

License

This project is licensed under the MIT License. See the LICENSE file for details.