Publications

Indoor-ALPS: An adaptive indoor location prediction system

Koehler C., Banovic N., Oakley I., Mankoff J., Dey A.K.

UbiComp 2014 - Proceedings of the 2014 ACM International Joint Conference on Pervasive and Ubiquitous Computing

2014

Conference Paper

Abstract:

Location prediction enables us to use a person’s mobility history to realize various applications such as efficient temperature control, opportunistic meeting support, and automated receptionists. Indoor location prediction is a challenging problem, particularly due to a high density of possible locations and short transition distances between these locations. In this paper we present Indoor-ALPS, an Adaptive Indoor Location Prediction System that uses temporal-spatial features to create individual daily models for the prediction of when a user will leave their current location (transition time) and the next location she will transition to. We tested Indoor-ALPS on the Augsburg Indoor Location Tracking Benchmark and compared our approach to the best performing temporal-spatial mobility prediction algorithm, Prediction by Partial Match (PPM). Our results show that Indoor-ALPS improves the temporal-spatial prediction accuracy over PPM for look-aheads up to 90 minutes by 6.2%, and for up to 30 minute look-aheads by 10.7%. These results demonstrate that Indoor-ALPS can be used to support a wide variety of indoor mobility prediction-based applications.

Information theoretical kernels for generative embeddings based on hidden markov models

Martins A.F.T., Bicego M., Murino V., Aguiar P.M.Q., Figueiredo M.A.T.

Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

2010

Conference Paper

Abstract:

Many approaches to learning classifiers for structured objects (e.g., shapes) use generative models in a Bayesian framework. However, state-of-the-art classifiers for vectorial data (e.g., support vector machines) are learned discriminatively. A generative embedding is a mapping from the object space into a fixed dimensional feature space, induced by a generative model which is usually learned from data. The fixed dimensionality of these feature spaces permits the use of state of the art discriminative machines based on vectorial representations, thus bringing together the best of the discriminative and generative paradigms. Using a generative embedding involves two steps: (i) defining and learning the generative model used to build the embedding; (ii) discriminatively learning a (maybe kernel) classifier on the adopted feature space. The literature on generative embeddings is essentially focused on step (i), usually adopting some standard off-the-shelf tool (e.g., an SVM with a linear or RBF kernel) for step (ii). In this paper, we follow a different route, by combining several Hidden Markov Models-based generative embeddings (including the classical Fisher score) with the recently proposed non-extensive information theoretic kernels. We test this methodology on a 2D shape recognition task, showing that the proposed method is competitive with the state-of-art.

Innovation by patients with rare diseases and chronic needs

Oliveira P., Zejnilovic L., Canhao H., Von Hippel E.

Orphanet Journal of Rare Diseases

2015

Article

Abstract:

We provide the first empirical exploration of disease-related innovation by patients and their caregivers. Our aims were to explore to what degree do patients develop innovative solutions; how many of these are unique developments; and do these solutions have positive perceived impact on the patients’ overall quality of life? In addition, we explored the factors associated with patient innovation development, and sharing of the solutions that the patients developed.

Integrating conceptual and procedural knowledge for middle-school math

Santos G., Howley I., Copenhaver B., Aleven V.

Frontiers in Artificial Intelligence and Applications

2009

Conference Paper

Abstract:

Cognitive Tutors have been shown to lead to impressive improvement in student learning in a range of domains, including middle school mathematics. Most Cognitive Tutors focus on providing learning support for tasks involving problem solving. They typically emphasize facilitating the acquisition of procedural knowledge, on the assumption that the acquisition of conceptual knowledge will be supported outside the tutor through classroom activities or an accompanying textbook. Many researchers agree that expertise in a domain involves not just procedural knowledge, but also conceptual knowledge, as well as tight interconnections between the two. In order to provide two types of knowledge instruction within the same Intelligent Tutoring System, we are now building a novel type of Cognitive Tutor to bridge the existing gap between teaching conceptual and procedural instruction with the same educational software.

Intent transfer in speech-to-speech machine translation

Anumanchipalli G.K., Oliveira L.C., Black A.W.

2012 IEEE Workshop on Spoken Language Technology, SLT 2012 - Proceedings

2012

Conference Paper

Abstract:

This paper presents an approach for transfer of speaker intent in speech-to-speech machine translation (S2SMT). Specifically, we describe techniques to retain the prominence patterns of the source language utterance through the translation pipeline and impose this information during speech synthesis in the target language. We first present an analysis of word focus across languages to motivate the problem of transfer. We then propose an approach for training an appropriate transfer function for intonation on a parallel speech corpus in the two languages within which the translation is carried out. We present our analysis and experiments on English↔Portuguese and English↔German language pairs and evaluate the proposed transformation techniques through objective measures.

Inter-application redundancy elimination in wireless sensor networks with compiler-assisted scheduling

Gupta V., Tovar E., Lakshmanan K., Rajkumar R.

7th IEEE International Symposium on Industrial Embedded Systems, SIES 2012 - Conference Proceedings

2012

Conference Paper

Abstract:

Most current-generation Wireless Sensor Network (WSN) nodes are equipped with multiple sensors of various types, and therefore support for multi-tasking and multiple concurrent applications is becoming increasingly common. This trend has been fostering the design of WSNs allowing several concurrent users to deploy applications with dissimilar requirements. In this paper, we extend the advantages of a holistic programming scheme by designing a novel compiler-assisted scheduling approach (called REIS) able to identify and eliminate redundancies across applications. To achieve this useful high-level optimization, we model each user application as a linear sequence of executable instructions. We show how well-known string-matching algorithms such as the Longest Common Subsequence (LCS) and the Shortest Common Super-sequence (SCS) can be used to produce an optimal merged monolithic sequence of the deployed applications that takes into account embedded scheduling information. We show that our approach can help in achieving about 60% average energy savings in processor usage compared to the normal execution of concurrent applications.

Interactive Optimal Teaching with Unknown Learners

Melo F.S. , Guerra C., Lopes M.

Proceedings of the Twenty-Seventh International Joint Conference on Artificial Intelligence (IJCAI-18)

2018

Conference Paper

Abstract:

This paper introduces a new approach for machine teaching that partly addresses the (unavoidable) mismatch between what the teacher assumes about the learning process of the student and the student’s actual process. We analyze several situations in which such mismatch takes place and weshow that, even in the simple case of a Bayesian Gaussian learner, the lack of knowledge regarding the student’s learning process significantly deteriorates the performance of machine teaching: while perfect knowledge of the student ensures that the target is learned after a finite number of samples, lack of knowledge thereof implies that the student will only learn asymptotically (i.e., after an infinite number of samples). We propose interactivity as a means to mitigate the impact of imperfect knowledge and show that, by using interactivity, we are able to attain significantly faster convergence, in the worst case. Finally, we discuss the implications of our results in single- and multi-student settings.

Investigating Utterance Level Representations for Detecting Intent from Acoustics

Rallabandi S., Karki b., Viegas C., Nyberg E., Black A.W.

Interspeech 2018

2018

Conference paper

Abstract:

Recognizing paralinguistic cues from speech has applications in varied domains of speech processing. In this paper we present approaches to identify the expressed intent from acoustics in the context of INTERSPEECH 2018 ComParE challenge. We have made submissions in three sub-challenges: prediction of 1) self-assessed affect and 2) atypical affect 3) Crying Sub challenge. Since emotion and intent are perceived at suprasegmental levels, we explore a variety of utterance level embeddings. The work includes experiments with both automatically derived as well as knowledge-inspired features that capture spoken intent at various acoustic levels. Incorporation of utterance level embeddings at the text level using an off the shelf phone decoder has also been investigated. The experiments impose constraints and manipulate the training procedure using heuristics from the data distribution. We conclude by presenting the preliminary results on the development and blind test sets.

Investigation of human identification using two-lead Electrocardiogram (ECG) signals

Ye C., Coimbra M.T., Kumar B.V.K.V.

IEEE 4th International Conference on Biometrics: Theory, Applications and Systems, BTAS 2010

2010

Conference Paper

Abstract:

In this paper, we investigate the applicability of Electrocardiogram (ECG) signals for human identification. Wavelet Transform (WT) and Independent Component Analysis (ICA) methods are applied to extract morphological features that appear to offer excellent discrimination among subjects. The proposed method is aimed at the two-lead ECG configuration that is routinely used in long-term continuous monitoring of heart activity. The information from the two ECG leads is fused to achieve improved subject identification. The proposed method was tested on three public ECG databases, namely, MIT-BIH Arrhythmias Database, Ml T-BIH Normal Sinus Rhythm Database and Long-Term ST Database, in order to evaluate the proposed subject identification method on normal ECG signals as well as ECG signals with arrhythmias. Excellent rank-1 recognition rates (as high as 99.6%) were achieved based on single heartbeats. The proposed method exhibits good identification accuracies not just with the normal ECG signals, but also in the presence of various arrhythmias. This work adds to the growing evidence that ECG signals can be useful for human identification.