Rigas Kouskouridas
Home - Research - Publications - Projects

June 2017: I have joined Scape, the most exciting start-up in London (according to many people in CVPR 2017).

Since September 2016, I am a computer vision / machine learning scientist at WIREWAX.

Before that I was a Postdoctoral Research Associate at the Computer Vision & Learning Lab at the Department of Electrical and Electronic Engineering of Imperial College London under the supervision of Assistant Professor Tae-Kyun Kim. I obtained my PhD in Computer Vision and Machine Learning in 2013 from the Department of Production Engineering and Management under the supervision of Associate Professor Antonios Gasteratos. I hold a MEng in Production Engineering and Management from the Democritus University of Thrace, Greece since 2006.

Research interests: Machine learning for vision with particular interest in object recognition, detection and pose estimation.

Email: rkouskou(at)gmail.com

August 2017The deadline for our ICCV 2017 workshop on "Recovering 6D Object Pose" is fast approaching!
July 2017ICCV 2017 paper accepted - Our work "Pose Guided RGBD Feature Learning for 3D Object Pose Estimation" has been accepted at ICCV 2017
June 2017Exiting times. Joining Scape, where I will focus on large scale 6D camera pose estimation.
September 2016Joining WIREWAX, the No.1 Interactive Video Technology, building cutting edge computer vision applications! Will try to keep this site upated.
July 2016We have uploaed our new dataset for the "Recovering 6D Object Pose and Predicting Next-Best-View Prediction in the Crowd" CVPR 2016 work
July 2016Recovering 6D Object Pose workshop paper deadline has been extended
July 2016Check out our new work "Siamese Regression Networks with Efficient mid-level Feature Extraction for 3D Object Pose Estimation"
July 2016The deadline for our ECCV 2016 workshop on "Recovering 6D Object Pose" is fast approaching!
March 2016Thanks to Wadim Kehl we have new corrected annotations for our Latent-Class Hough Forests dataset
March 2016CVPR 2016 paper - "Recovering 6D Object Pose and Predicting Next-Best-View Prediction in the Crowd" - Source code and dataset coming soon
March 2016New work - "Iterative Hough Forest with Histogram of Control Points for 6 DoF Object Registration from Depth Images"
February 2016Live Demonstration - "Hands in Action" - Imperial Festival (Research Zone) , 7-8 May 2016, Imperial College London, UK
January 2016Matlab Tutorial - Part of the Machine Learning for Computer Vision course
December 2015New work - "6D Object Detection and Next-Best-View Prediction in the Crowd" - Visit project page
December 2015Our dataset for evaluation of illumination invariant feature detectors and descriptors migrated to a new host
December 2015Live demonstration of our work on real-time 6DoF object pose estimation at the first international workshop on Recovering 6D Object Pose held in conjuction with ICCV 2015
November 2015I have designed a project page that lists state of the art methods and datasets available to further help drive research in the field of object recognition, detection and 6D pose estimation
September 2015ACPR 2015 paper accepted - Our work "Video-based Object Recognition with Weakly Supervised Object Localization" has been accepted at ACPR 2015
June 2015I am co-organising the ICCV 2015 workshop on Recovering 6D Object Pose. With Andreas Doumanoglou we are responsible for organising the Multi-Object Pose Estimation Challenge
May 2015Invited Talk - "Robot Vision in the The Self-Conscious Robot" - Pint of Science - The Bolton, 326 Earls Court Road, London, SW5 9BQ
May 2015Live Demonstration "6-Dimensional Robot Vision" - Imperial College London - Imperial Festival
November 2014"Random Forests and Applications - Object Detection and Pose Estimation" - Imperial College London EE course Machine Learning for Computer Vision - Download Slides
October 2014GTA Machine Learning for Computer Vision Department of Electrical & Electronic Engineering - Imperial College London - Course website - Office Hours Tue 11am - 12noon, Room 1003
August 2014ECCV Object Dataset Released - Our challenging object dataset is now available. More than 5K images of 6 different objects under altering viewpoints with background clutter, foreground occlusions and multi-instance objects
June 2014ECCV 2014 paper accepted - Our work "Latent-Class Hough Forests for 3D Object Detection and Pose Estimation" has been accepted at ECCV 2014