News

2015-11-20

TUT Doctoral School Funding

Junsheng Fu received two year funding for his PhD thesis work "Large Scale 3D" from the TUT Doctoral School.

The competition was extremely tough (115 applications and 25 positions) so congratulations Junsheng!


2015-11-18

FiDiPro Professor Jiri Matas joins the Vision Group

Professor Jiri Matas becomes a FiDiPro professor in the Vision Group from Jan 2016 to Dec 2019. He will mentor and supervise senior and junior researchers and collaborate with the local companies.

Tekes FiDiPro program provides funding to recruit top international researchers to Finnish research teams. These researchers, often distinguished professors in their fields, bring scientific expertise and research practices that boost the teams toward the top international level.

The Vision Group of Tampere University of Technology (Dept. of Signal Processing) received funding for a FiDiPro professor and his research team for four years starting January 2016. The project is a consortium project together with the Center for Machine Vision Research (CMV).

FiDiPro Professor Jiri Matas from the Center for Machine Perception, Czech Technical University, has published many pioneering and seminal works of computer vision, has been the program chair of the top international conferences (CVPR, ICCV and ECCV) and is the Associate Editor-in-Chief of the most prestigious journal of computer vision and machine learning (IEEE-PAMI). Moreover, his works have received numerous best paper awards.

Associate Professor Joni Kamarainen: "Professor Matas is a kind of researcher who can leave his permanent mark to any group he visits and who will certainly be remembered. He is creative, ambitious and extremely hard working and demanding, both to his students and colleagues. His impact to our field - measured in the number of citations - matches to the rest of us in the department, so I expect that he will help to lead our scientific topics and research practices toward those of the very best groups in the world. We are privileged!"

This Tekes funded FiDiPro will also benefit the local companies who are project partners:

  • Digile Oy
  • Fingersoft Oy
  • Haltian Oy
  • Intel Finland Oy
  • IntoPalo Oy
  • NOMO Technologies Oy
  • Visy Oy
For more information contact:
  • Joni Kämäräinen, Tampere university of Technology
  • Janne Heikkilä, University of Oulu
  • Jiri Matas, Czech Technical University


2015-09-25

Ekaterina Riabchenko's PhD Defense

MSc Ekaterina Riabchenko successfully defended her PhD thesis "Generative Part-Based Gabor Object Detector".

Riabchenko PhD thesis: PDF. The thesis is related to the Gabor Object Detector project (www) where Dr. Riabchenko developed a novel Gabor feature based generative method for learning and detecting visual classes and their poses.

The opponents in the thesis examination committee were Professor Ales Leonardis from University of Birmingham, UK, and Dr. Jorma Laaksonen from Aalto University, Finland.

The final statement by Prof. Ales Leonardis


2015-09-24

Vision Group participates the TUT Flagship project "Robotics and Intelligent Machines"

The four department units of Tampere University of Technology started an ambitious research project on autonomous robots and heavy machinery. The project is funded by TUT and Academy of Finland. The goal is a completely autonomous work site of multiple co-operating robots and machines.

Tampere University of Technology is profiling in teaching and research on robotics and, in particular, heavy working machinery. Robotics research is expensive, challenging and multidisciplinary research field, but it has enormous future potential. This has been identified in the EU Horizon 2020 programme which aims to bring European research and industry to the front line of "robotization". In Tampere region and Finland there are several big players who benefit from the national top research and TUT has long research and industrial collaboration tradition in the field of heavy machinery.

The Flagship project was started in 2015 by the four active departments with existing research in the robotics:

  • Intelligent Hydraulics and Automation (Professors Kalevi Huhtala and Reza Ghabcheloo)
  • Mechanical Engineering and Industrial Systems (Professor Kari T Koskinen)
  • Signal Processing (Professors Joni Kämäräinen, Ari Visa and Heikki Huttunen)
  • Automation Science and Engineering (Professor Risto Ritala who is the project leader)
  • The goal of the project is a robotized environment (work site), where robots and heavy machinery autonomously operate and collaborate and in the case of failures or other "surprises" collaborative re-plan and repair machinery and operation with the help of available (re-usable and movable) robot platform. The system is based on multiple robust sensors and, in particular, computer vision and machine learning.

    See the example video of the system: https://www.youtube.com/watch?v=30U1Pe0-k_E Schematic picture of the TUT-RIM worksite for robotic experiments


    2015-03-09

    A paper accepted to CVPR 2015

    Our CVPR 2015 paper introduces novel visual similarity networks with applications in unsupervised image alignment.

    The TUT capital gain funded project "BigData" resulted to a novel concept of "visual similarity graphs" which was developed in the Vision Group of the Dept. of Signal Processing. The method automatically constructs a graph (network) where similar images have strong links between them and the network represents visual similarities in natural way allowing efficient graph algorithms in large scale visual processing. The results will be presented in the IEEE Conf. on Computer Vision and Pattern Recognition (CVPR) in Boston, Massachusetts, USA in June 2015 in

    • Unsupervised Visual Alignment with Similarity Graphs (F. Shokrollahi Yancheshmeh, K. Chen, J.-K. Kämäräinen), In IEEE Conf. on Computer Vision and Pattern Recognition (CVPR), 2015.
    CVPR is the premier annual Computer Vision event. According to Google Scholar Metrics, CVPR is the top publication venue in the field of computer vision and pattern recognition. CVPR also ranks 7th in the category of Engineering & Computer Science, and it is the highest-ranked venue in Computer Science.
    Visual similarity graph of motorbikes
    Visual similarity graph of motorbikes


    2015-03-03

    Hardware Donation from NVidia

    Vision Group received hardware donation fron NVidia corporation. The powerful NVidia Tesla K40 GPU card will support our deep neural networks research.

    Convolutional deep neural networks have become an important part of computational intelligence and machine learning recently. An interesting computational property of the existing libraries is that they can be significantly boosted with powerful GPU hardware and even commodity GPUs beat cutting edge CPUs. Therefore for research groups it is important to have access to high end GPUs in their research. NVidia corporation donated the Vision Group one of the high end GPUs, NVidia Tesla K40, for research purposes.

    Researcher Yuan Liu is happy user of the new powerful hardware
    Researcher Yuan Liu is the happy user of the new powerful NVidia donated hardware


    2015-01-23

    D2I Funding for Deep Neural Networks Research

    Vision Group participates to Data to Intelligence Program (D2I) of the Digile ICT SHOK (Finland Strategic Centre for Science, Technology and Innovation in ICT). Vision Group focus is on efficient deep neural network algorithms for visual data mining.

    Data to Intelligence (D2I) program is a large research project gathering together the main Finnish players of ICT industry, technology and science. The project consists of many topics that share multiple ICT topics, but one of the central themes is automatic analysis of visual content (images and videos). The recent breakthroughs using deep learning techniques are investigated in the 2015 work packages of D2I in which TUT Vision Group participates.


    2014-11-19

    TUT Doctoral School Funding

    Fatemeh Shokrollahi Yancheshmeh received four year funding for her PhD project "Rich Geometry and Appearance Discover for Large Scale Image Analysis" from TUT Doctoral School.

    The competition was extremely tough (112 applications and 25 positions) so congratulations Fatemeh! The project will support our Academy of Finland project related to large scale image analysis.


    2014-05-27

    FishView Project Kickoff

    EU funded FishView - Assessing Fish Passages by the Use of Robotic Fish Sensor and Enchanced Digital Imaging - project started.

    The EU funded FishView (Assessing Fish Passages by the Use of Robotic Fish Sensor and Enchanced Digital Imaging) project combines knowledge from robotics (Prof. Maarja Kruusmaa (project coordinator), Centre for Biorobotics, Tallinn University of Technology, Estonia), signal processing (Prof. Joni Kamarainen, Department of Signal Processing, Tampere University of Technology, Finland), and ecohydraulic engineering (Dr. Jeff Tuhtan, SJE Ecohydraulic Engineering, GmbH). The project goal is to develop a novel fish robot based sensor to evaluate flows and conditions of rivers and fish passes and help to assess and develop better fish passages. Project description in the EU BONUS database: FishView.

    FishView project coordinator Prof. Maarja Kruusmaa and the fish robot sensor prototype
    FishView project coordinator Prof. Maarja Kruusmaa and the fish robot sensor prototype


    2014-05-22

    Jukka Lankinen's PhD

    MSc Jukka Lankinen successfully defended his PhD thesis "Local Features in Image and Video Processing - Object Class Matching and Video Shot Detection".

    Tykkälä PhD thesis: PDF. The thesis is related to the Camera2DVD project (www) where he developed novel approaches for object alingment, video processing and skimming of home videos.

    The opponents in the thesis examination were Dr. Krystian Mikolajczyk from University of Surrey, UK, and Dr. Esa Rahtu, University of Oulu, Finland.