Low Light Image Enhancement

🚀 Zero-DCE for low-light image enhancement

Implement Zero-Reference Deep Curve Estimation for enhancing low-light images.

🚀 Quick Start

This project focuses on implementing Zero-Reference Deep Curve Estimation (Zero-DCE) for low-light image enhancement. It formulates the enhancement task as estimating an image - specific tonal curve using a deep neural network.

✨ Features

• Zero-Reference Approach: Zero-DCE doesn't require any input/output image pairs during training. It uses a set of non - reference loss functions to implicitly measure the enhancement quality and guide network training.
• Tonal Curve Estimation: A lightweight deep network, DCE - Net, is trained to estimate pixel - wise and high - order tonal curves for dynamic range adjustment of a given image.
• Contrast Preservation: The curve estimation process maintains the range of the enhanced image and preserves the contrast of neighboring pixels.

📚 Documentation

General Information

• Original Author: Soumik Rakshit
• Date created: 2021/09/18
• HF Contribution: Harveen Singh Chadha
• Dataset: LOL Dataset
• Spaces Demo: Click here

Technical Details

Zero-Reference Deep Curve Estimation (Zero-DCE) formulates low-light image enhancement as the task of estimating an image - specific tonal curve with a deep neural network. In this example, a lightweight deep network, DCE - Net, is trained to estimate pixel - wise and high - order tonal curves for dynamic range adjustment of a given image.

Zero-DCE takes a low - light image as input and produces high - order tonal curves as its output. These curves are then used for pixel - wise adjustment on the dynamic range of the input to obtain an enhanced image. The curve estimation process is done in such a way that it maintains the range of the enhanced image and preserves the contrast of neighboring pixels. This curve estimation is inspired by curves adjustment used in photo - editing software such as Adobe Photoshop where users can adjust points throughout an image’s tonal range.

Zero-DCE is appealing because of its relaxed assumptions with regard to reference images: it does not require any input/output image pairs during training. This is achieved through a set of carefully formulated non - reference loss functions, which implicitly measure the enhancement quality and guide the training of the network.

Sample Images

📄 License

This project is licensed under the Apache - 2.0 license.