Guest editorial - Special issue on robust recognition methods for multimodal interaction
Guardado en:
| Autor principal: | |
|---|---|
| Otros Autores: | , |
| Formato: | Capítulo de libro |
| Lenguaje: | Inglés |
| Publicado: |
Elsevier B.V.
2014
|
| Acceso en línea: | Registro en Scopus DOI Handle Registro en la Biblioteca Digital |
| Aporte de: | Registro referencial: Solicitar el recurso aquí |
| LEADER | 11954caa a22003497a 4500 | ||
|---|---|---|---|
| 001 | PAPER-23899 | ||
| 003 | AR-BaUEN | ||
| 005 | 20230518205540.0 | ||
| 008 | 190411s2014 xx ||||fo|||| 00| 0 eng|d | ||
| 024 | 7 | |2 scopus |a 2-s2.0-84893691145 | |
| 040 | |a Scopus |b spa |c AR-BaUEN |d AR-BaUEN | ||
| 100 | 1 | |a Gomez, L. | |
| 245 | 1 | 0 | |a Guest editorial - Special issue on robust recognition methods for multimodal interaction |
| 260 | |b Elsevier B.V. |c 2014 | ||
| 506 | |2 openaire |e Política editorial | ||
| 536 | |a Detalles de la financiación: Universidad de Buenos Aires | ||
| 536 | |a Detalles de la financiación: National Institutes of Natural Sciences | ||
| 536 | |a Detalles de la financiación: Purdue University | ||
| 536 | |a Detalles de la financiación: Luis Gomez Guest Editors lgomez@ctim.es CTIM: Image Technology Center, Electronic Engineering and Automatic Control Department, University of Las Palmas de Gran Canaria, Campus de Tafira, 35017 Las Palmas, Spain CTIM: Image Technology Center Electronic Engineering and Automatic Control Department University of Las Palmas de Gran Canaria Campus de Tafira 35017 Las Palmas Spain J.P. Wachs jpwachs@purdue.edu Intelligent Systems and Assistive Technologies Laboratory (ISAT), School of Industrial Engineering, Purdue University, 315 North Grant Street, West Lafayette, 47907 IN, United States Intelligent Systems and Assistive Technologies Laboratory (ISAT) School of Industrial Engineering Purdue University 315 North Grant Street West Lafayette 47907 IN United States Julio Jacobo-Berlles jacobo@dc.uba.ar Robotics and Embedded Systems Laboratory, Computer Science Department, Faculty of Exact and Natural Sciences, Ciudad Universitaria, University of Buenos Aires, Argentina Robotics and Embedded Systems Laboratory Computer Science Department Faculty of Exact and Natural Sciences Ciudad Universitaria University of Buenos Aires Argentina This Special Issue of the Elsevier Journal of Pattern Recognition Letters entitled “ Robust Recognition Methods for Multimodal Interaction ” presents the ongoing and active research in the Pattern Recognition discipline applied to multimodal interfaces. The main advantage of multimodal systems is that they are intuitive, natural and allow rich expressiveness of particular interest in the special issue are real-time applications of multimodal interfaces to healthcare, assistive technologies, robotics, entertainment and context-aware systems. Multimodal interfaces allow us to benefit from streams of rich, expressive real-time sensed data that can otherwise be overwhelming. In this context, one of the main challenges still to be tackled is the design of robust and effective pattern recognition algorithms that can not only extract significant features to make sense of these data, but whose modalities can also self-adjust their integration so that ambiguity in one modality is always resolved by the certainty of the remaining ones. Twelve papers constitute this Special Issue. Some of them are extended versions of selected papers presented at the 17th Iberoamerican Congress on Pattern Recognition (CIARP), held at Buenos Aires (Argentina) between September 3–6, 2012. The rest of articles are external submissions from the call for papers. At least two reviewers assessed the quality of the papers, and those meeting the top standards were sent to a second round of reviews. The guest editors hope that the selected papers will provide the readers with interesting examples of current research on the most outstanding theoretical frameworks in the pattern recognition context applied to multimodal interfaces as well as in the challenging field of applications through practical and efficient algorithms. Based on the problems presented, we classified them into the following categories: review paper, human–computer interaction, human-action recognition, biometry, computer vision, music information retrieval and speckle reduction. This Special Issue opens with a paper entitled “ Multimodal Interaction: A Review ” by M. Turk discussing pioneering works, progress, methods and challenges in the multimodal interaction field. This paper discussed the motivation behind multimodal interaction as the driving force to provide a more natural, powerful, and compelling interactive experiences to the user. Turk addresses the fundamentals of interacting through different modalities, such as speech, touch, vision and gestures, and the advantages and limitations of each. This paper also brings-up a major research question related to whether the modalities should be integrated early or late in the process, supporting the claims by recent evidence in biological sensory theory. By the end of the paper, a number of challenges and opportunities are described including the number of modalities to be used, and the way on which those should be combined in practical applications. The reviews and discussion presented are placed in the context of the broad computer vision area. The area of human–computer interaction recognition was represented by two papers. The first one, entitled, “ Context-based Hand Gesture Recognition for the Operating Room ” by M. Jacob and J.P. Wachs, addresses the challenge of developing a sterile interface for browsing medical images in the operating room that relies uniquely on gestural communication. The main challenge addressed is to avoid gesture recognition false alarms while keeping high recognition accuracy. For this, contextual information is utilized. The most important scientific contribution of this work is a methodology to automatically determine the optimal parameters required for Hidden Markov Models (HMM) for continuous recognition of hand gestures. Gestures are segmented and classified automatically using gesture spotting networks (GSN). These networks include both independently trained HMM models (one for each hand gesture class) together with a threshold model (used to represent unintentional movements). The second article devoted to human–computer interaction is “ Towards a Real-Time Interface between a Biomimetic Model of Sensorimotor Cortex and a Robotic Arm ”, by S. Dura-Bernal, G. Chadderdon, S. Neymotin, J. Francis and W. Lytton. The authors present the first steps for the integration of a biomimetic spiking neuronal model of motor learning with a robotic arm. The biomimetic model was used to drive a simple kinematic two-joint virtual arm in a motor task requiring trial-and-error convergence on a single target. The area of human action recognition was represented by the following two papers. The first one, entitled, “ Efficient Descriptor Tree Growing for Fast Action Recognition ”, by S. Ubalde, N. Goussies and M. Mejail. The presented method is based on an efficient computation of the Instance-to-Class distance (I2C), which can be applied if a clever organization of training data is previously performed. An algorithm for organizing data in such a way is described. The result is an easy-to-train classifier which achieves state-of-the-art performance and reasonable computation times when working with large training databases. The second article, “ On the Improvement of Human Action Recognition from Depth Map Sequences using Space-Time Occupancy Patterns” , by authors A. Vieira, E. Nascimento, G. Oliveira, Z. Liu and M. Campos, proposes a new visual representation for 3D action recognition from sequences of depth maps. Space and time axes are divided into multiple segments to define a 4D space-time grid. The occupancy values of the cells of this grid are used to define a high dimensional feature vector that preserves spatial and temporal contextual information between space-time cells. On the subject of biometry, four papers were selected. The first one entitled “ Robust Gender Recognition by Exploiting Facial Attributes Dependencies ”, by J. Belkios-Calfa, J. Buenaposada and L. Baumela studies the dependencies among gender, age and pose facial attributes. Here, the existence of these dependencies is confirmed experimentally and this proves that they can be exploited to improve the performance and robustness of face-based gender classifiers. In the article entitled “ Face Recognition on Partially Occluded Images using Compressed Sensing ” by A. Morelli, S. Padovani, M. Tepper and J. Jacobo-Berlles, authors deal with the very challenging problem of face recognition for partially occluded images. The proposal is based on previous well-known work by Wright et al., but including remarkable modifications to detect and discard occluded areas of the face. Next manuscript dealing with biometric interfaces is “Efficient Software Attack to Multimodal Biometric Systems and its Application to Face and Iris Fusion” by M. Gómez-Barrero, J. Galbally and J. Fierrez. In this article, authors present an innovative system that explores the vulnerabilities of multi-modal biometric systems to software-based attacks. Its performance is tested against a biometric system based on iris and face identification. The article, “ Fused Intra-bimodal Face Verification Approach based on Scale-invariant Feature Transform and a Vocabulary Tree ” by authors C.M. Travieso, M. del Pozo-Baños and J.B. Alonso closes the Biometric topic. This is a paper that evaluates recognition of people based on face and head images acquired from the thermal IR and the visible range of the electromagnetic spectrum. The evaluation is made separately and also making data fusion. The latter is done by comparing a score function (sum and product) and a decision function (OR function and weight function). It uses SIFT descriptors, and a vocabulary tree to store them in a database. This Special Issue closes with the following three manuscripts: The first of them, entitled, “ Line Detection in Images Showing Significant Lens Distortion and Application to Distortion Correction ”, by M. Aleman-Flores, L. Alvarez, L. Gomez and D. Santana-Cedres presents a novel and robust method to automatically detect lines in images with the aim to later correct the radial distortion caused by the lens. The method proposed extends the usual Hough representation by introducing a new parameter, the lens distortion, in such a way that the new search space becomes a three-dimensional space accounting for the orientation, the distance to the origin and also the radial distortion. Music Information Retrieval was the area of the paper, “ Query by Humming: Automatically Building the Database from Music Recordings ” by M. Rocamora, P. Cancela and A. Pardo. This paper proposes a method for extracting | ||
| 593 | |a Electronic Engineering and Automatic Control Department, Campus de Tafira, University of Las Palmas de Gran Canaria, 35017 Las Palmas, Spain | ||
| 593 | |a Intelligent Systems and Assistive Technologies Laboratory (ISAT), School of Industrial Engineering, Purdue University, 315 North Grant Street, West Lafayette, IN 47907, United States | ||
| 593 | |a Computer Science Department, Ciudad Universitaria, University of Buenos Aires, Argentina | ||
| 700 | 1 | |a Wachs, J.P. | |
| 700 | 1 | |a Jacobo-Berlles, J. | |
| 773 | 0 | |d Elsevier B.V., 2014 |g v. 36 |h pp. 187-188 |k n. 1 |p Pattern Recogn. Lett. |x 01678655 |t Pattern Recognition Letters | |
| 856 | 4 | 1 | |u https://www.scopus.com/inward/record.uri?eid=2-s2.0-84893691145&doi=10.1016%2fj.patrec.2013.09.010&partnerID=40&md5=ee9c19be93e7d2d98630bf93dadadafa |y Registro en Scopus |
| 856 | 4 | 0 | |u https://doi.org/10.1016/j.patrec.2013.09.010 |y DOI |
| 856 | 4 | 0 | |u https://hdl.handle.net/20.500.12110/paper_01678655_v36_n1_p187_Gomez |y Handle |
| 856 | 4 | 0 | |u https://bibliotecadigital.exactas.uba.ar/collection/paper/document/paper_01678655_v36_n1_p187_Gomez |y Registro en la Biblioteca Digital |
| 961 | |a paper_01678655_v36_n1_p187_Gomez |b paper |c PE | ||
| 962 | |a info:eu-repo/semantics/article |a info:ar-repo/semantics/artículo |b info:eu-repo/semantics/publishedVersion | ||
| 999 | |c 84852 | ||