📞 Digital Arrest Detector

An AI-powered tool to detect digital fraud, scams, and deepfakes across multiple channels including Text, Audio, and Video.
Built for real-time prevention, alerts, and awareness against modern cyber scams.

---

🚀 Features

• Text Scam Detection

  • NLP-based classification (legit vs scam)
  • Keyword spotting for risky terms
  • Sentiment analysis for urgency/fear/threat

• Audio Scam Detection

  • Speech-to-text transcription (supports en-in, hi, gu)
  • ML-based scam probability scoring
  • Placeholder for deepfake/voice spoofing detection

• Video Deepfake Detection

  • Keras-based deepfake detection model (Deepfakes_detection_model.keras)
  • Classifies uploaded videos as Likely Real / Deepfake

---

🛠 Tech Stack

• Backend: Python (Streamlit)
• ML/NLP: Scikit-learn, TF-IDF, Keras
• Audio Processing: Vosk, SoundFile, Wave
• Video Processing: OpenCV, TensorFlow/Keras
• Deployment Ready: Streamlit (demo) / React or Vue (future)

---

📂 Project Structure

Digital-arrest-detector/
│── app.py                # Main app (Streamlit UI)
│── models/               # ML models (text, tfidf, deepfake, etc.)
│── data/                 # Sample data for training
│── train_text.py         # Script to train text scam detection model
│── temp_video.mp4        # Example video file for deepfake detection
│── temp.wav              # Example audio file for scam detection
│── requirements.txt      # Dependencies
│── README.md             # Project documentation

---

⚙️ Installation

1. Clone the repo:

   git clone https://github.com/VishvaNarkar/Digital-arrest-detector.git
   cd Digital-arrest-detector

2. Run the app:

   streamlit run app.py

---

🖥 Usage

• Text Analysis → Paste SMS/Email/Chat text to detect scams
• Audio Analysis → Upload call recording (wav/mp3) → Transcription + Scam detection
• Video Analysis → Upload video (mp4/avi) → Deepfake detection

---

📌 Future Roadmap

• 🔲 Integrate real-time call/email blocking
• 🔲 Advanced multi-language NLP models
• 🔲 Deploy full-stack version (FastAPI + React/Vue)

---

📜 License

MIT License – free to use and modify with attribution.