Related Experiment Video
Updated: Dec 3, 2025
03:31
Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications
Published on: December 15, 2023
855
Unsupervised Dark-Channel Attention-Guided CycleGAN for Single-Image Dehazing
Jiahao Chen1, Chong Wu1, Hu Chen1,2
1National Key Laboratory of Fundamental Science on Synthetic Vision, Chengdu 610000, China.
Sensors (Basel, Switzerland)
|October 29, 2020
Summary
This study introduces an unsupervised attention-based cycle generative adversarial network (GAN) for single-image dehazing. The novel method effectively removes haze while preserving image details, outperforming existing techniques in diverse conditions.
Area of Science:
- Computer Vision
- Artificial Intelligence
- Image Processing
Background:
- Single-image dehazing remains a challenge, particularly in complex scenes with varying haze concentrations.
- Traditional generative adversarial networks (GANs) can alter haze-free regions and struggle with detail preservation.
Purpose of the Study:
- To develop an unsupervised attention-based cycle GAN for effective single-image dehazing.
- To improve haze removal by preserving original image details and addressing varying haze levels.
Main Methods:
- Proposed an unsupervised attention-based cycle generative adversarial network (GAN).
- Incorporated an attention mechanism to selectively dehaze different image regions.
- Utilized training-enhanced dark channels as attention maps, combining prior algorithms and deep learning.
Main Results:
- The method successfully dehazed images without requiring paired datasets.
- High-resolution image generation was achieved.
- Demonstrated superior performance compared to previous dehazing algorithms across various scenarios.
Conclusions:
- The proposed attention-based GAN offers a significant advancement in single-image dehazing.
- Effective for processing images with heavy haze, mist, and even haze-free images in complex scenes.
- Preserves image details while accurately addressing haze concentrations.
Related Concept Videos
01:20Deconvolution
434
Deconvolution, also known as inverse filtering, is the process of extracting the impulse response from known input and output signals. This technique is vital in scenarios where the system's characteristics are unknown, and they must be inferred from the observable signals.
Deconvolution involves several mathematical techniques to derive the impulse response. One common approach is polynomial division. In this method, the input and output sequences are treated as coefficients of...
Deconvolution involves several mathematical techniques to derive the impulse response. One common approach is polynomial division. In this method, the input and output sequences are treated as coefficients of...
434
02:16Light Acquisition
9.1K
In order to produce glucose, plants need to capture sufficient light energy. Many modern plants have evolved leaves specialized for light acquisition. Leaves can be only millimeters in width or tens of meters wide, depending on the environment. Due to competition for sunlight, evolution has driven the evolution of increasingly larger leaves and taller plants, to avoid shading by their neighbors with contaminant elaboration of root architecture and mechanisms to transport water and nutrients.
9.1K