QuantaLux is a hybrid deep learning–quantum computing project designed to convert low-light or nighttime images into realistic daytime visuals. By integrating a classical CycleGAN architecture with quantum circuits using Qiskit, QuantaLux aims to enhance scalability, convergence speed, and visual quality in image-to-image translation tasks.
This project explores the use of Quantum Computing techniques combined with CycleGAN to transform nighttime images into daylight equivalents. The system is designed to assist in:
- Enhancing remote sensing and surveillance under low-light conditions
- Improving image visibility for autonomous systems in night-time environments
- Exploring quantum acceleration for real-world deep learning tasks
- ⚛️ Quantum-Enhanced CycleGAN: Integrates quantum feature mapping into the standard CycleGAN architecture.
- 🌃 Night-to-Day Image Translation: Converts nighttime scenes into daytime images with improved lighting and detail.
- 🧮 Qiskit Integration: Uses quantum circuits to explore high-dimensional representations of image features.
- 📊 Scalable Architecture: Modular design with configurable layers, loss functions, and training modes.
- 🎯 Future Scope: Extendable to other low-light enhancement or generative tasks.