Densely Connected Pyramid Dehazing Network (2018-03-22T00:00:00.000000Z)

TL;DR

A new end-to-end single image dehazing method, called Densely Connected Pyramid Dehazing Network (DCPDN), which can jointly learn the transmission map, atmospheric light and dehazed all together by directly embedding the atmospheric scattering model into the network.

Abstract

We propose a new end-to-end single image dehazing method, called Densely Connected Pyramid Dehazing Network (DCPDN), which can jointly learn the transmission map, atmospheric light and dehazing all together. The end-to-end learning is achieved by directly embedding the atmospheric scattering model into the network, thereby ensuring that the proposed method strictly follows the physics-driven scattering model for dehazing. Inspired by the dense network that can maximize the information flow along features from different levels, we propose a new edge-preserving densely connected encoder-decoder structure with multi-level pyramid pooling module for estimating the transmission map. This network is optimized using a newly introduced edge-preserving loss function. To further incorporate the mutual structural information between the estimated transmission map and the dehazed result, we propose a joint-discriminator based on generative adversarial network framework to decide whether the corresponding dehazed image and the estimated transmission map are real or fake. An ablation study is conducted to demonstrate the effectiveness of each module evaluated at both estimated transmission map and dehazed result. Extensive experiments demonstrate that the proposed method achieves significant improvements over the state-of-the-art methods. Code and dataset is made available at: https://github.com/hezhangsprinter/DCPDN

Authors

Vishal M. Patel

23 papers

He Zhang

5 papers

TL;DR

Abstract

Authors

References58 items

Jointly Optimize Data Augmentation and Network Training: Adversarial Data Augmentation in Human Pose Estimation

Gated Fusion Network for Single Image Dehazing

Context Encoding for Semantic Segmentation

Photographic Text-to-Image Synthesis with a Hierarchically-Nested Adversarial Network

Residual Dense Network for Image Super-Resolution

Density-Aware Single Image De-raining Using a Multi-stream Dense Network

Face Synthesis from Visual Attributes via Sketch using Conditional VAEs and GANs

RESIDE: A Benchmark for Single Image Dehazing

Stacked Conditional Generative Adversarial Networks for Jointly Learning Shadow Detection and Shadow Removal

A Generative Adversarial Approach for Zero-Shot Learning from Noisy Texts

High-Resolution Image Synthesis and Semantic Manipulation with Conditional GANs

High-Quality Facial Photo-Sketch Synthesis Using Multi-Adversarial Networks

HazeRD: An outdoor scene dataset and benchmark for single image dehazing

Joint Transmission Map Estimation and Dehazing Using Deep Networks

Generating High-Quality Crowd Density Maps Using Contextual Pyramid CNNs

An All-in-One Network for Dehazing and Beyond

DenseNet for dense flow

Air-light estimation using haze-lines

Image De-Raining Using a Conditional Generative Adversarial Network

NIPS 2016 Tutorial: Generative Adversarial Networks

StackGAN: Text to Photo-Realistic Image Synthesis with Stacked Generative Adversarial Networks

DeMoN: Depth and Motion Network for Learning Monocular Stereo

Pyramid Scene Parsing Network

The One Hundred Layers Tiramisu: Fully Convolutional DenseNets for Semantic Segmentation

Image-to-Image Translation with Conditional Adversarial Networks

Single Image Dehazing via Multi-scale Convolutional Neural Networks

Photo-Realistic Single Image Super-Resolution Using a Generative Adversarial Network

Densely Connected Convolutional Networks

A Two-Streamed Network for Estimating Fine-Scaled Depth Maps from Single RGB Images

Non-local Image Dehazing

Coupled Generative Adversarial Networks

Generative Adversarial Text to Image Synthesis

Perceptual Losses for Real-Time Style Transfer and Super-Resolution

DehazeNet: An End-to-End System for Single Image Haze Removal

Stage-wise Training: An Improved Feature Learning Strategy for Deep Models

Multimodal deep learning for robust RGB-D object recognition

A Fast Single Image Haze Removal Algorithm Using Color Attenuation Prior

SUN RGB-D: A RGB-D scene understanding benchmark suite

Simultaneous video defogging and stereo reconstruction

U-Net: Convolutional Networks for Biomedical Image Segmentation

Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift

Dehazing Using Color-Lines

Very Deep Convolutional Networks for Large-Scale Image Recognition

High-Resolution Stereo Datasets with Subpixel-Accurate Ground Truth

Investigating Haze-Relevant Features in a Learning Framework for Image Dehazing

Spatial Pyramid Pooling in Deep Convolutional Networks for Visual Recognition

Automatic recovery of the atmospheric light in hazy images

Visibility Restoration of Single Hazy Images Captured in Real-World Weather Conditions

Efficient Image Dehazing with Boundary Constraint and Contextual Regularization

Visualizing and Understanding Convolutional Networks

Single Image Dehazing by Multi-Scale Fusion

Indoor Segmentation and Support Inference from RGBD Images

Factorizing Scene Albedo and Depth from a Single Foggy Image

Deep photo: model-based photograph enhancement and viewing

Single image dehazing

Visibility in bad weather from a single image

GENERATIVE ADVERSARIAL NETS

Single Image Haze Removal Using Dark Channel Prior

Field of Study

Journal Information

Name

Page