🎙 STTSSTTSSTTSSTTS: Speech-to-Text & Sentiment Analysis

🚀 STTSSTTSSTTSSTTSS is a tool for transcribing audio files and analyzing sentiment using Yandex SpeechKit, RemBERT trained on KazSAnDRA dataset created by ISSAI, and Streamlit.

📌 Features

• ✅ Transcribe audio files (WAV, MP3, FLAC, etc.)
• ✅ Sentiment analysis (positive, neutral, negative)
• ✅ Supports Kazakh and Russian languages
• ✅ User-friendly UI with Streamlit
• ✅ Leverages Yandex Cloud API as a submodule

---

📥 Installation & Setup

🔧 Prerequisites

• Python 3.12 (Ensure Python 3.12 is installed)
• FFmpeg (Required for audio processing)

  sudo apt install ffmpeg  # Linux
  brew install ffmpeg      # macOS

• Git (For cloning the repository and initializing submodules)

---

🚀 Clone the Repository

Since this project uses Yandex Cloud API as a submodule, use:

git clone --recurse-submodules https://github.com/tvran/Forte-stt.git
cd Forte-stt

If you have already cloned the repo without submodules, initialize it manually:

git submodule update --init --recursive

---

🛠 Generate gRPC Client Interface

To use Yandex SpeechKit, you need to generate the gRPC client interface.

1️⃣ Install grpcio-tools

pip install grpcio-tools

2️⃣ Run the following command inside the Forte-STT directory:

python3 -m grpc_tools.protoc -I cloudapi -I cloudapi/third_party/googleapis \
  --python_out=output \
  --grpc_python_out=output \
  cloudapi/google/api/http.proto \
  cloudapi/google/api/annotations.proto \
  cloudapi/yandex/cloud/api/operation.proto \
  cloudapi/google/rpc/status.proto \
  cloudapi/yandex/cloud/operation/operation.proto \
  cloudapi/yandex/cloud/validation.proto \
  cloudapi/yandex/cloud/ai/stt/v3/stt_service.proto \
  cloudapi/yandex/cloud/ai/stt/v3/stt.proto

This will generate necessary Python files in output/:

• stt_pb2.py
• stt_pb2_grpc.py
• stt_service_pb2.py
• stt_service_pb2_grpc.py

---

📦 Install Dependencies

Activate a virtual environment (recommended):

python3 -m venv venv
source venv/bin/activate  # On Windows: venv\Scripts\activate

Then, install the required dependencies:

pip install -r requirements.txt

---

🚀 Running the Project

Start the Streamlit UI

streamlit run main.py

If deploying on a server, use:

streamlit run main.py --server.port 8501

---

🚀 Setting up API Keys

To use Yandex SpeechKit and Hugging Face Transformers, you need to store API keys securely.

1️⃣ Create a .env file in the root of the project

2️⃣ Add your API keys inside .env:

# Yandex SpeechKit API Key
YANDEX_API_KEY=your_yandex_api_key_here

# Yandex Object Storage Keys
ACCESS_KEY=your_access_key_here
SECRET_KEY=your_secret_key_here

# Hugging Face Token (for sentiment analysis)
HF_TOKEN=your_huggingface_token_here

---

📂 Project Structure

Forte-stt/
│── output/                  # Audio processing & recognition logic
│   ├── adjust_audio.py       # Converts audio to 16kHz PCM
│   ├── load_file.py          # Uploads to Yandex Cloud Storage
│   ├── recognize.py          # Handles Yandex SpeechKit transcription
│   ├── stt_pb2.py            # gRPC-generated file
│   ├── stt_service_pb2.py    # gRPC-generated file
│── cloudapi/                 # Yandex Cloud API (submodule)
│── main.py                   # Streamlit UI
│── requirements.txt          # Python dependencies
│── README.md                 # Documentation

---

🛠 Technologies Used

• Python 3.12
• Streamlit – UI for audio processing
• Yandex SpeechKit – Speech-to-Text processing
• Hugging Face Transformers – Sentiment analysis
• FFmpeg – Audio conversion
• gRPC – Communication with Yandex API

---

📞 Contact

👤 Turan Nurgozhin
📧 Email: turannurgozhin@gmail.com
🔗 LinkedIn: https://www.linkedin.com/in/turan-nurgozhin-81931428b/
🚀 GitHub: github.com/tvran