Academy
Search
Close this search box.

Publications

A model checking framework for linear time invariant switching systems using structural systems analysis

Ramos G., Pequito S., Aguiar A.P., Ramos J., Kar S.
2013 51st Annual Allerton Conference on Communication, Control, and Computing, Allerton 2013
2013
Conference Paper
Abstract:
This paper introduces the concept of structural hybrid systems to address, as a particular case, the model checking problem of switching (possible large scale) linear time invariant systems. Within the proposed setup, we provide necessary conditions to ensure properties such as controllability, at each time. We show that such model checking controllability properties can be implemented using efficient algorithms (with polynomial complexity). An example, based on the IEEE 5-bus power system, is presented which illustrates our model checking and design methodologies.

A model of primate photoreceptors

Goncalves H.R., Correia M.V.
BIOSIGNALS 2010 - Proceedings of the 3rd International Conference on Bio-inpsired Systems and Signal Processing, Proceedings
2010
Conference Paper
Abstract:
As experimental research reveals the biological mechanisms behind the processing done by the retina, complete models of the retina become more and more possible. This paper presents a temporal model of primate photoreceptors inspired by the mechanisms discovered in other species. It implements light adaptation based on pigment bleaching and biochemical reactions. The simulation provides similar results to experiments made in impulse, contrast and sensitivity response curves of primate cones and rods.

A No-Reference Quality Metric for Retinal Vessel Tree Segmentation

Galdran A., Costa P., Bria A., Araújo T., Mendonça A.M., Campilho A.
MICCAI 2018: Medical Image Computing and Computer Assisted Intervention
2018
Conference Paper
Abstract:
Due to inevitable differences between the data used for training modern CAD systems and the data encountered when they are deployed in clinical scenarios, the ability to automatically assess the quality of predictions when no expert annotation is available can be critical. In this paper, we propose a new method for quality assessment of retinal vessel tree segmentations in the absence of a reference ground-truth. For this, we artificially degrade expert-annotated vessel map segmentations and then train a CNN to predict the similarity between the degraded images and their corresponding ground-truths. This similarity can be interpreted as a proxy to the quality of a segmentation. The proposed model can produce a visually meaningful quality score, effectively predicting the quality of a vessel tree segmentation in the absence of a manually segmented reference. We further demonstrate the usefulness of our approach by applying it to automatically find a threshold for soft probabilistic segmentations on a per-image basis. For an independent state-of-the-art unsupervised vessel segmentation technique, the thresholds selected by our approach lead to statistically significant improvements in F1-score (+2.67%)(+2.67%) and Matthews Correlation Coefficient (+3.11%3.11%) over the thresholds derived from ROC analysis on the training set. The score is also shown to correlate strongly with F1 and MCC when a reference is available.

A Pixel-Wise Distance Regression Approach for Joint Retinal Optical Disc and Fovea Detection

Meyer M.I., Galdran A., Mendonça A.M., Campilho A.
MICCAI 2018: Medical Image Computing and Computer Assisted Intervention
2018
Conference Paper
Abstract:
This paper introduces a novel strategy for the task of simultaneously locating two key anatomical landmarks in retinal images of the eye fundus, namely the optic disc and the fovea. For that, instead of attempting to classify each pixel as belonging to the background, the optic disc, or the fovea center, which would lead to a highly class-imbalanced setting, the problem is reformulated as a pixelwise regression task. The regressed quantity consists of the distance from the closest landmark of interest. A Fully-Convolutional Deep Neural Network is optimized to predict this distance for each image location, implicitly casting the problem into a per-pixel Multi-Task Learning approach by which a globally consistent distribution of distances across the entire image can be learned. Once trained, the two minimal distances predicted by the model are selected as the locations of the optic disc and the fovea. The joint learning of every pixel position relative to the optic disc and the fovea favors an automatic understanding of the overall anatomical distribution. This results in an effective technique that can detect both locations simultaneously, as opposed to previous methods that handle both tasks separately. Comprehensive experimental results on a large public dataset validate the proposed approach.

A platform for Operator-driven Network Virtualization

Nogueira J., Melo M., Carapinha J., Sargento S.
EUROCON 2011 - International Conference on Computer as a Tool - Joint with Conftele 2011
2011
Conference Paper
Abstract:
Network virtualization enables the coexistence of multiple networks, running different protocols, in an infrastructure-independent way. With that goal in mind, this paper presents a Platform for Operator-driven Network Virtualization that builds virtual networks through a user-friendly interface, integrating virtual network mapping and creation, discovery, monitoring, and management functionalities. Besides the developed functionalities, this platform contains novel mechanisms for network discovery and mapping: a novel dynamic distributed discovery algorithm of both physical and virtual nodes, and a heuristic algorithm for mapping virtual resources in the physical infrastructure, that supports the heterogeneity of networks, with respect to both links and nodes. The platform and its features were implemented and evaluated in different scenarios. The obtained results show the scalability and feasibility of the proposed mechanisms and functionalities in a single platform for network virtualization control and management.

A population model for predicting human mobility

Baumann P., Dey A., Koehler C., Santini S.
UbiComp and ISWC 2015 - Proceedings of the 2015 ACM International Joint Conference on Pervasive and Ubiquitous Computing and the Proceedings of the 2015 ACM International Symposium on Wearable Computers
2015
Conference Paper
Abstract:
We present a population model for predicting users’ next relevant place. Our findings demonstrate the robustness of our population model in terms of the difference in accuracy with respect to users’ individual models, which is for more than 90% of the users less than 3.5 percentage points.

A Quest for Inspiration: How Users Create and Reuse PINs

Casimiro M., Segel J., Li L., Wang Y., Cranor L F.
WAY2020
2020
Article
Abstract:
Personal Identification Numbers (PINs), required to authenticate on a multitude of devices, are ubiquitous nowadays. To increase the security and safety of their assets, users are advised to create unique PINs for a lot of accounts they possess. Considering the multiple accounts users hold, remembering a myriad of PINs is often burdensome for users. As a consequence, we suspect users tend to trade-off security for memorability, due to the fear of forgetting their PINs, thus reusing them. To test this hypothesis we conducted a study on MTurk that asked participants about their PIN creation and reuse behaviors. Our results show that users draw inspiration from important dates to create their PINs and that PIN reuse is common practice, even between high and low valued accounts. Participants justify this behavior stating they reuse PINs for convenience and ease of remembrance.

  • Year

    -

  • Publication type

Contacts

Facebook-f Instagram X-twitter Linkedin-in Youtube