Computational
imaging group

World leader in image and video denoising and enhancement.

Tampere university

The perfect place for science and education

Tampere University home page

Computational Imaging Group

Is a part of The Computing Sciences Unit, Faculty of Information Technology and Communication Sciences, Tampere University, Tampere, Finland. Founded in 1998. The group leader is Professor Karen Egiazarian.
The mission of Computational Imaging Group is a theoretical and applied research different from the traditional image processing with a special emphasis on applied mathematics, optics, human perception. Applications are in consumer electronics, mobile imaging, microscopy, and remote sensing.

Faculty of Information Technology
and Communication Sciences

Research Areas

Our research area is digital image processing. Under this area our main research interests are:

Sparse image and video processing in transform domain

The sparsity concepts claims that fragments of natural image can be represented by a linear combination of small number of atoms (basis functions). It is a flexible instrument for various imaging problems such as denoising, deblurring, demosaicing, super-resolution, compressed sensing, and other inverse problems in imaging.

Coherent phase imaging and light field phase coding

Phase imaging is a unique instrument to study details of internal structure of transparent or semitransparent specimens. In modern technology and science phase and wavefield imaging are popular and well established technique for high-accuracy measuring, recording and reconstructing 2D and 3D objects. The areas of applications are varying from astronomy and engineering to medicine and biology.

Image visual quality assessment

Full-reference and no-reference image visual quality metrics, verification of quality metrics, human visual system, blind noise parameters estimation

Image and video lossy compression

DCT based compression, fractal based compression, partitions schemes in lossy compression, multichannel image compression, noisy image compression, automatic selection of compression parameters

JUST NOW WE WORK ON

Deep learning neural network based image denoising,
Image visual quality assessment,
Coherent phase imaging

Join our research!

Our experiences

We are interested in joint applications to scientific projects and grants
Results of our investigations were/are state-of-the-art in the following research areas

Image and signal denoising

BM3D filter last decade is state-of-the-art in image denoising

Image visual quality assessment

TID2013 is widely used to evaluate image visual quality metrics

Lossy image compression

ADCT coder several years was state-of-the-art in lossy image compression

Blind noise parameter estimation

Our methods provide best precision of estimation of noise variance and noise spectrum

Our Projects

Some of scientific projects and grants of our group

CIWiL

Computational imaging without lens

H2020-TWINN-2015 HOLO

Boosting the scientific excellence and innovation capacity in digital holographic microscopy

SPARSE

Sparse complex domain wave field imaging

SpaRTaN

Sparse Representations and Compressed Sensing Training Network

MacSeNet

Machine Sensing Training Network

INTERFERO

Advanced image processing for interferometric and noninterferometric phase and wave field reconstruction

FLEXISP

End-to-end image processing pipeline for Flexible ISP SOW

Phase imaging

Deep CNN computational lens for phase imaging

Meet The Team

Current members of Computational Imaging Group

Karen Egiazarian

PhD, DTech

Professor at Tampere University


Vladimir Katkovnik

PhD, DTech

Professor, Senior Researcher at Tampere University


Mykola Ponomarenko

PhD, DTech

Senior Research Fellow at Tampere University, Professor at National Aerospace University, Ukraine

Igor Shevkunov

PhD

Senior Researcher at Tampere University


Samuel Pinilla

PhD

Postdoctoral Research Fellow at Tampere University

Peter Kocsis

PhD student

PhD student at Tampere University

Sheyda Ghanbaralizadeh Bahnemiri

PhD student

PhD student at Tampere University

Seyyed Reza Miri Rostami

PhD student

PhD student at Tampere University

Wenzhu Xing

PhD student

PhD student at Tampere University

Shushik Avagyan

PhD student

PhD student at Tampere University

Oleg Ieremeiev

PhD student

PhD student at Tampere University, Senior Tutor at National Aerospace University, Ukraine

Oleksii Rubel

PhD student

PhD student at Tampere University, Senior Tutor at National Aerospace University, Ukraine

Veronica Cazac

PhD student

PhD student at Tampere University,Young Researcher of the Recording Media and Photonics group of the Materials for Photovoltaic and Photonics Laboratory (IAP)


Our collaborators:

Vladimir Lukin

PhD, DTech

Visiting Researcher at Tampere University, Professor at National Aerospace University, Ukraine

Sergey Abramov

PhD

Visiting Researcher at Tampere University, Associate Professor at National Aerospace University, Ukraine

Alexander Totsky

PhD, DTech

Visiting Researcher at Tampere University, Professor at National Aerospace University, Ukraine

Nikolay Petrov

PhD

Senior Researcher, Associate Professor, Head of the Digital and Display Holography Laboratory, Department of Photonics and Optical Information Technologies ITMO University, Russia

Elena Achimova

PhD

Head of the Recording Media and Photonics group of the Materials for Photovoltaic and Photonics Laboratory at the Institute of Applied Physics

Vladimir Abaskin

PhD

Leading researcher of the Recording Media and Photonics group of the Materials for Photovoltaic and Photonics Laboratory at the Institute of Applied Physics

Alexei Meshalkin

Researcher of the Recording Media and Photonics group of the Materials for Photovoltaic and Photonics Laboratory (IAP)


Our Skills

DIGITAL IMAGE PROCESSING
97%
MACHINE LEARNING
85%
DATA COMPRESSION
85%
SOFTWARE DESIGN
82%

Our History

Transforms and Spectral Techniques Group was established.

Part of SPAG (Signal Processing Algorithm Group) elected as an Academy of Finland Centre of Excellence for 2000-2005.

Part of SPAG (Signal Processing Algorithm Group) elected as an Academy of Finland Centre of Excellence for 2006-2011.

Changed name from "Transforms and Spectral Techniques" to "Computational Imaging".

The group joins Signal processing research community (SPRC) at Tampere University of Technology.

Tampere University was established as a merger between the University of Tampere and Tampere University of Technology.

Faqs

Please contact to Professor Karen Egiazarian, room TF415 (or see his contact data in the section "Contacts" on the web page).
Send your complete CV to Professor Karen Egiazarian.
Ask one of co-authors of the publication (e.g. through Research Gate). Also you may check section "Publications" in this web page.

How we work? :)

Typical flow-chart of our investigations

1

IDEA

2

STUDY

3

PLAN

4

DESIGN

5

VERIFICATION

6

PUBLICATION

<

Selected Publications

A small repository (full texts, codes, web pages) of selected publications of Computational Imaging Group
If you need a full list of our publications press the button:    Full list on Google Scholar

Pages of selected publications

Sparse image and video processing in transform domain   Open in a new window
Coherent phase imaging, light field phase coding, lensless optics   Open in a new window
Image visual quality assessment   Open in a new window
Blind noise parameter estimation   Open in a new window
Image and video lossy compression   Open in a new window

Fun Facts

Some statistic data, based on Google Scholar

NUMBER OF PUBLICATIONS
OF OUR GROUP
NUMBER OF CITATIONS
FOR ALL PUBLICATIONS
NUMBER OF CITATIONS
FOR THE MOST CITED PUBLICATION
H-INDEX
OF OUR GROUP

Latest Publications

Some of the recent publications can be found bellow

Xing W., Egiazarian K.
Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2021, pp. 3507-3516

End-to-end learning for joint image demosaicing, denoising and super-resolution

Image denoising, demosaicing and super-resolution are key problems of image restoration well studied in the recent decades. Often, in practice, one has to solve these problems simultaneously. A problem of finding a joint solution of the multiple image restoration tasks just begun to attract an increased attention of researchers. In this paper, we propose an end-to-end solution for the joint demosaicing, denoising and super-resolution based on a specially designed deep convolutional neural network (CNN).

Read More

Our collaboration

Our collaboration includes, but is not limited to:

Institute of Applied Physics of the Academy of Sciences of Moldova

Coordinator of the HOLO project

National Aerospace University of Ukraine

Number of joint scientific investigations

University of Stuttgart

Participant of the HOLO project

Noiseless imaging

Number of joint projects

Get in Touch

Where we are and how to contact us

Contact Info

Tampere University
The Computing Sciences Unit
P.O. Box 553
FI-33101 Tampere
Finland

Phone: +358 408415663
Professor Karen Egiazarian
Send Message