Deep task-driven video denoising

The main contribution of this research is two folds. First, this research work explores the vast domain of video denoising, analyze challenges in designing video denoising algorithm, and study previously successful state-of-the-art video denoising techniques. Secondly, this research aims to optimize...

Full description

Saved in:
Bibliographic Details
Main Author: Kurniadi, Daniel
Other Authors: Wen Bihan
Format: Final Year Project
Language:English
Published: Nanyang Technological University 2020
Subjects:
Online Access:https://hdl.handle.net/10356/138731
Tags: Add Tag
No Tags, Be the first to tag this record!
Institution: Nanyang Technological University
Language: English
Description
Summary:The main contribution of this research is two folds. First, this research work explores the vast domain of video denoising, analyze challenges in designing video denoising algorithm, and study previously successful state-of-the-art video denoising techniques. Secondly, this research aims to optimize video denoising algorithm when the result is supplied for high-level task behind it. We propose a method that features high-level information guided video denoising that capable of achieving comparable result with the state-of-the-art denoising while preserving semantic-aware details for high-level vision tasks.