IPE Seminar

Reduced Exposure and Fast Imaging in Industrial CT Using Noise2Noise Denoising Method

by Matin Dehghani

→ Europe/Berlin
Teams

Teams

https://teams.microsoft.com/l/meetup-join/19%3ameeting_ZjI2ODI3N2MtOGJjZC00ZjdhLWIyYWUtYjBkNmZiYmMxNmMw%40thread.v2/0?context=%7b%22Tid%22%3a%224f5eec75-46fd-43f8-8d24-62bebd9771e5%22%2c%22Oid%22%3a%22cff71760-dfd4-4016-a9ec-07ff82a9f34e%22%7d Besprechungs-ID: 342 213 912 692 Kennung: JZwrAe
Description

Internship presentation of Matin Dehghani on the topic : Reduced Exposure and Fast Imaging in Industrial CT Using Noise2Noise Denoising Method

This study explores using the Noise2Noise framework to enhance X-ray CT scans in industrial settings, focusing on reducing scan times.

Unlike traditional methods, Noise2Noise trains on noisy, low-exposure datasets without needing a noise-free ground truth. The report compares Noise2Noise with other reconstruction algorithms like TV regularization, examining its potential to expedite scans and improve image quality.

 

Lecture language: English