News and Events PDF Print E-mail

Seminar 16 October 2017 - Mubarak Shah - 14:00 Aula Magna

Title: Solving Semantic Segmentation: Precision Matrix, Knowledge-Based Rules and Generator Adversarial Network (GAN)

Speaker: Mubarak Shah

Abstract : Figure-ground separation is a land-mark problem in visual perception, which has fascinated many scientists for centuries. In computer vision, edge detection and region segmentation of an image have been two grand challenges for understanding of an image in terms of objects and contextual surroundings, and shapes and appearances of objects.  Generic Segmentation of an image involves grouping pixels, which are perceptually similar. However, in Semantic Segmentation the aim is to assign a semantic label to each pixel in the image. Even though semantic segmentation can be achieved by simply applying classifiers (which are trained using supervised learning), to each pixel or a region in the image, the results may not be desirable due to the fact that general context information beyond the simple smoothness is not considered. In this talk, I will start with briefly presenting two supervised approaches to address this problem.  First, I will discuss an approach to discover interactions between labels and regions using a sparse estimation of precision matrix, which is the inverse of covariance matrix of data obtained by graphical lasso. In this context, we find a graph over labels as well as regions in the image which encodes significant interactions and also it is able to capture the long-distance associations. Second, I will introduce a knowledge-based method to incorporate dependencies among regions in the image during inference. High level knowledge rules - such as co-occurrence, spatial relations and mutual exclusivity - are extracted from training data and transformed into constraints in Integer Programming formulation.
A difficulty which most supervised semantic segmentation approaches are confronted with is lack of enough training data, particularly in deep learning methods which have become enormously popular recently. Annotated data should be at the pixel-level (i.e., each pixel of training images must be annotated), which is highly expensive to achieve. To address this limitation, next I will present a semi supervised learning approach to exploit the plentiful amount of available unlabeled as well as synthetic images generated via Generative Adversarial Networks (GAN). Furthermore, I will discuss an extension of the model to use additional weakly labeled data to solve the problem in a weakly supervised manner. The basic idea here is by providing these fake data from the Generator and the competition between real/fake data (discriminator/generator networks), true samples are encouraged to be close in the feature space. Therefore, the model learns more discriminative features, which lead to better classification results for semantic segmentation.

Bio: Dr. Mubarak Shah, the UCF Trustee Chair Professor, is the founding director of Center for Research in Computer Visions at University of Central Florida (UCF). He is a co-author of five books (Motion-Based Recognition (1997); Video Registration (2003); Automated Multi-Camera Surveillance: Algorithms and Practice (2008); Modeling, Simulation and Visual Analysis of Crowds (2013); and Robust Subspace Estimation Using Low-Rank Optimization (2014), all by Springer.  He has published extensively on topics related to visual surveillance, tracking, human activity and action recognition, object detection and categorization, shape from shading, geo registration, visual crowd analysis, etc. Dr. Shah is a fellow of IEEE, IAPR, AAAS and SPIE.  He has been ACM and IEEE Distinguished Visitor Program speaker and is often invited to present seminars, tutorials and invited talks all over the world. He received Pegasus award in 2006; University Distinguished Research Award in 2017, 2012 and 2005; Faculty Excellence in Mentoring Doctoral Students in 2016, Scholarship of Teaching and Learning award in 2011; Teaching Incentive Program award in 1995 and 2003; Research Incentive Award in 2003, 2009 and 2012; the Harris Corporation Engineering Achievement Award in 1999; the TOKTEN awards from UNDP in 1995, 1997, and 2000; 2009 IEEE Outstanding Engineering Educator Award in 1997; an honorable mention for the ICCV 2005 Where Am I? Challenge Problem; 2013 NGA Best Research Poster Presentation, 2nd place in Grand Challenge at the ACM Multimedia 2013 conference; and runner up for the best paper award in ACM Multimedia Conference in 2005 and 2010.m I? Challenge Problem; 2013 NGA Best Research Poster Presentation, 2nd place in Grand Challenge at the ACM Multimedia 2013 conference; and runner up for the best paper award in ACM Multimedia Conference in 2005 and 2010.


TRADR and Alcor in Piombino for the European Robotics League (ERL) Event

On September 17 2017, the TRADR and Alcor teams gave a live demo in Piombino, for the European Robotics League event. The robots Romeo and Delta successfully showed how they can autonomously patrol an arena with fences and barricades. We were pleased to present the robots and describe how they are able to achieve autonomous navigation, mapping and patrolling in the crowded Piazza Bovio.

TRADR robots patrolling



Seminar Tuesday 12 September 2017 - Sean Ryan Fanello - 11:00 Aula Magna

Title: Low Compute and Fully Parallel Computer Vision with HashMatch

Speaker: Sean Ryan Fanello

Abstract: Depth cameras are becoming key tools for computer vision tasks ranging from hand, body or object tracking, 3D reconstruction and simultaneous localization and mapping. Almost all these tasks need to solve a *tracking* problem i.e. each new frame of depth and image data is correlated to the previous, and this temporal information allows for faithful pose and/or geometry reconstruction over time. However, this reliance on temporal information, makes tracking problems hard to solve when using sensors running at 30fps, due to susceptibility to high frame-to-frame scene motions and artifacts such as motion blur. Using high speed depth cameras would greatly simplify these problems and make them more tractable, but despite significant research efforts, no existing high framerate and high quality depth algorithm, and hence camera exists.
In this talk I am going to demonstrate a 3D capture system for high speed and high quality depth estimation, and show its advantages in a variety of computer vision tasks. Our hardware and software depth pipeline can run at 1.1msec with modern GPUs and readily procurable camera and illumination components.

Bio: Sean Ryan Fanello has received his Bachelor and Master degrees from Sapienza where he worked on 3D Reconstruction and Gesture Recognition. He received his PhD in Robotics from Italian Institute of Technology where he designed and implemented the iCub perception system. In particular, he worked on algorithms for 3D estimation, hand-eye calibration, egomotion estimation, action recognition and object recognition. After his PhD, he spent 3 years at Microsoft Research. His research is mainly focusing on the intersection among Machine Learning, Computer Vision and Natural User Interfaces. At Microsoft he has developed new technologies for 3D estimation that have been deployed on the Microsoft Hololens headset. He was one of the leading members of important projects like Holoportation where they showed the first 3D telepresence system with HD quality. Holoportation, enables real-time ''teleportation'' of people from anywhere in the world allowing them to communicate fully in 3D. Recently he is a founding team member of perceptiveIO. At perceptiveIO he is developing novel 3D sensing capabilities, natural user interfaces and computer vision applications.

TRADR Summer School 2017 - Human-Robot Teaming

As partners of the EU-FP7-ICT-Project TRADR, we are glad to announce the TRADR Summer School 2017 on Human-Robot Teaming. The Summer School will take place from Monday 21st to Friday 25th August in Utrecht, The Netherlands.

This event aims at providing the participants with a full overview and hands-on experience on real robotic systems in disaster response scenario mainly focusing on aspects related to human-robot teaming.  Teaming and collaboration aspects that occur between humans and robots will also be presented in healthcare, education and space exploration doman applications.

Preliminary information about speakers, dates and submission guidelines are available here or on


The TRADR EU Project on Euronews

The TRADR EU project is on Euronews. The video report shows the simulated disaster scene we used in the last project review and some shots from the mission in Amatrice with Vigili del Fuoco. The troupe of journalists came to visit us at the DIAG and interviewed our Team. More details on this page.





The ALCOR Team at the Edison Pulse Day

The ALCOR Team has been invited to talk at the Edison Pulse Day. The work accomplished within the TRADR EU Project and the mission in Amatrice will be presented as an inspirational and successful history in the context. This Italian event will host some of the most innovative Italian startups. Further details at this page.



TRADR Year 3 Review in Montelibretti - Work Package 4 Obtained an Excellent Evaluation

The TRADR Year 3 Review took place at the Scuola Di Formazione Operativa of Vigili del Fuoco (Italian Firefighters) in Montelibretti (Italy), on Tuesday March 7 and Wednesday March 8.
Work Package 4, under ALCOR team responsability, obtained an excellent evaluation. During the demo, we showed:
• our new path planner nicely driving the robots among wrecked cars, stones and barriers in a rescue environment
• two robots patrolling in the gallery

You can find further details about our new patrolling algorithm at this page.








© 2017 Alcor
Joomla! is Free Software released under the GNU General Public License.