Summary of Deep Variational Bayesian Modeling Of Haze Degradation Process, by Eun Woo Im et al.
Deep Variational Bayesian Modeling of Haze Degradation Process
by Eun Woo Im, Junsung Shin, Sungyong Baik, Tae Hyun Kim
First submitted to arxiv on: 4 Dec 2024
Categories
- Main: Computer Vision and Pattern Recognition (cs.CV)
- Secondary: Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary The proposed variational Bayesian framework for single-image dehazing tackles the neglect of transmission and atmospheric light factors in recent neural-network-based approaches. By introducing a transmission map as a latent variable, parameterized by a separate neural network, the framework encourages cooperation between dehazing and transmission networks through a new objective function based on a physical haze degradation model. This joint training boosts performance, and the approach can be seamlessly incorporated with existing dehazing networks. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary Imagine you’re trying to remove fog from a photo. Most recent attempts have ignored important factors like how much light is reaching your eyes from far away, and how bright everything looks in different parts of the scene. To fix this, researchers developed a new way to “unfog” pictures using two special types of neural networks: one for removing haze and another for estimating how much light is being blocked. This combination leads to better results than previous methods, and it can even be used with other dehazing approaches to make them work better too. |
Keywords
» Artificial intelligence » Neural network » Objective function
