stt-research-scripts

Python scripts and experiments used for my research paper on speech-to-text AI models. This repository includes model evaluations, accuracy benchmarks, and transcript quality comparisons using various open-source and cloud-based STT APIs.

> Place your audio file in the same directory and update `audio_path` in the scripts.

Setup and run Whisper

Follow these steps to set up and run the Whisper transcription script:

1. Use Python 3.10

Make sure you have Python 3.10 installed on your machine.

⚠️ Whisper requires Python 3.10 due to dependency compatibility. Using other versions (like 3.11 or 3.12) may cause installation issues or runtime errors.

2. Create a virtual environment

python3.10 -m venv whisper-venv

3. Activate the virtual environment

source whisper-venv/bin/activate

4. Install dependencies

pip install openai-whisper

5. Run the transcription script

python3 stt-whisper-model.py

6. Deactivate the virtual environment

deactivate

Setup and run Vosk

1. Download the English Vosk Model (lightweight)

wget https://alphacephei.com/vosk/models/vosk-model-small-en-us-0.15.zip
unzip vosk-model-small-en-us-0.15.zip
mv vosk-model-small-en-us-0.15 vosk-model

This downloads and extracts the model to a folder named vosk-model which the script expects.

2. Create a virtual environment

python3.10 -m venv vosk-venv

3. Activate the virtual environment

source vosk-venv/bin/activate

4. Install dependencies

pip install vosk

5. Run the transcription script

python3 stt-vosk-model.py path/to/your/audio.wav

Replace path/to/your/audio.wav with the path to your WAV file (mono, 16kHz).

6. Output files

After running, two files will be generated in the script directory:

transcription.json — full transcription plus word-level timestamps transcription.txt — full transcription as plain text

7. Deactivate the virtual environment

deactivate

Setup and run AssemblyAI

1. Create a virtual environment

python3.10 -m venv assemblyai-venv

2. Activate the virtual environment

source assemblyai-venv/bin/activate

3. Install dependencies

pip install requests, dotenv

4. Run the transcription script

python3 stt-assemblyai-model.py

5. Deactivate the virtual environment

deactivate