Lowlight enhancement - Mirnet open - source model - Free improvement of the clarity and details of low

Lowlight Enhance Mirnet

Developed by keras-io

MIRNet is a fully convolutional neural network specifically designed for low-light image enhancement tasks, capable of fusing multi-scale contextual information while preserving high-resolution details.

Image Enhancement #Low-light enhancement #Multi-scale fusion #Fully convolutional architecture

Downloads 86

Release Time : 3/2/2022

Model Overview

This model reconstructs high-quality images from low-light input images through deep learning techniques, applicable in fields such as photography, security surveillance, medical imaging, and remote sensing.

Model Features

Multi-scale feature fusion

Simultaneously captures local and global contextual information through unique architectural design

Detail preservation

Effectively maintains high-frequency details and texture features during enhancement

End-to-end training

Adopts end-to-end training to directly learn the mapping relationship from low-light to normal exposure

Model Capabilities

Low-light image enhancement

Image quality restoration

Detail enhancement

Use Cases

Photography enhancement

Night photography enhancement

Improves photo quality taken by smartphones or cameras under low-light conditions

Visible noise reduction and detail enhancement in comparison

Security surveillance

Surveillance video enhancement

Improves visibility of low-light surveillance footage

Increases accuracy in face and object recognition

🚀 tf-keras

This library provides a model for low - light image enhancement using MIRNet, offering solutions for image restoration in various fields.

🚀 Quick Start

This README provides a detailed introduction to the MIRNet model for low - light image enhancement, including model description, dataset information, training procedures, and more.

✨ Features

The MIRNet model can recover high - quality image content from degraded low - light images.
It uses a fully - convolutional architecture to learn enriched features, combining multi - scale contextual information and preserving high - resolution spatial details.

📚 Documentation

🔍 Model description

This repo contains the model and the notebook Low - light image enhancement using MIRNet.

Full credits go to Soumik Rakshit.

Reproduced by Vu Minh Chien with a slight change on hyperparameters.

With the goal of recovering high - quality image content from its degraded version, image restoration enjoys numerous applications, such as photography, security, medical imaging, and remote sensing. The MIRNet model for low - light image enhancement is a fully - convolutional architecture that learns an enriched set of features that combines contextual information from multiple scales, while simultaneously preserving the high - resolution spatial details.

📦 Dataset

The [LoL Dataset](https://drive.google.com/uc?id=1DdGIJ4PZPlF2ikl8mNM9V - PdVxVLbQi6) has been created for low - light image enhancement. It provides 485 images for training and 15 for testing. Each image pair in the dataset consists of a low - light input image and its corresponding well - exposed reference image.

🔧 Training procedure

Training hyperparameters

The following hyperparameters were used during training:

learning_rate: 1e - 04
train_batch_size: 8
seed: 42
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e - 08
lr_scheduler_type: ReduceLROnPlateau
num_epochs: 50

Training results

The results are shown in TensorBoard (Training metrics).

View Model Demo

Model Demo

View Model Plot

![Model Image](./model.png)

Featured Recommended AI Models

Empowering the Future, Your AI Solution Knowledge Base

English 简体中文繁體中文にほんご