Sentiment Analysis Flask App

This Flask application performs:

• Text sentiment analysis using the CardiffNLP Twitter RoBERTa model.
• Audio sentiment analysis by transcribing uploaded audio and analyzing the text sentiment.
• Tone analysis of customer feedback.
• Text-to-speech generation using ElevenLabs API.

---

Setup

1. Clone the repository and navigate to the project directory.

2. Create and activate a Python virtual environment:

python3 -m venv venv
source venv/bin/activate

3. Install dependencies:

pip install -r requirements.txt

4. Make sure ffmpeg is installed on your system:

ffmpeg -version

If not installed, install it via your package manager, e.g.,

sudo apt install ffmpeg

---

Usage

Run the Flask app:

python app.py

---

API Endpoints

• POST /analyze/text - Sentiment analysis for raw text.
• POST /analyze/audio - Sentiment analysis for audio files.
• POST /generate/audio - Generate speech audio from text using ElevenLabs API.

---

Important Notes

• For audio files, the app converts audio to WAV format before transcription.
• When using ElevenLabs text-to-speech output, you may need to convert the audio sample rate and channels for compatibility, e.g.:

ffmpeg -i elevenlabs_output.wav -ar 16000 -ac 1 -c:a pcm_s16le elevenlabs_output1.wav

This converts the audio to 16 kHz, mono channel, and PCM signed 16-bit little-endian format, often required for ASR or other audio processing tools.

---

CORS

CORS is enabled in the app for local Swagger UI and cross-origin requests.

---

Configuration

Set your ElevenLabs API key in the app (ELEVENLABS_API_KEY).