INESC-ID

About: INESC-ID is a nonprofit organization based out in Lisbon, Portugal. It is known for research contribution in the topics: Field-programmable gate array & Control theory. The organization has 932 authors who have published 2618 publications receiving 37658 citations.

Battle of the DJs: an HCI perspective of Traditional, Virtual, Hybrid and Multitouch DJing

Abstract: The DJ culture uses a gesture lexicon strongly rooted in the traditional setup of turntables and a mixer. As novel tools are introduced in the DJ community, this lexicon is adapted to the features they provide. In particular, multitouch technologies can oer a new syntax while still supporting the old lexicon, which is desired by DJs. We present a classication of DJ tools, from an interaction point of view, that divides the previous work into Traditional, Virtual and Hybrid setups. Moreover, we present a multitouch tabletop application, developed with a group of DJ consultants to ensure an adequate implementation of the traditional gesture lexicon. To conclude, we conduct an expert evaluation, with ten DJ users in which we compare the three DJ setups with our prototype. The study revealed that our proposal suits expectations of Club/Radio-DJs, but fails against the mental model of Scratch-DJs, due to the lack of haptic feedback to represent the record’s physical rotation. Furthermore, tests show that our multitouch DJ setup, reduces task duration when compared with Virtual setups.

Controlling complexity in part-of-speech induction

Abstract: We consider the problem of fully unsupervised learning of grammatical (part-of-speech) categories from unlabeled text. The standard maximum-likelihood hidden Markov model for this task performs poorly, because of its weak inductive bias and large model capacity. We address this problem by refining the model and modifying the learning objective to control its capacity via parametric and non-parametric constraints. Our approach enforces word-category association sparsity, adds morphological and orthographic features, and eliminates hard-to-estimate parameters for rare words. We develop an efficient learning algorithm that is not much more computationally intensive than standard training. We also provide an open-source implementation of the algorithm. Our experiments on five diverse languages (Bulgarian, Danish, English, Portuguese, Spanish) achieve significant improvements compared with previous methods for the same task.

A real-time low-cost marker-based multiple camera tracking solution for virtual reality applications

Abstract: One of the reasons of how virtual reality applications penetrate industrial production chains slowly is because of significant investment costs to purchase adequate supporting hardware. As a consequence such applications are only available to major companies and fail to benefit the production processes of small and medium enterprises. In this article, we introduce PTrack, a real-time, low-cost, marker-based multiple camera tracking solution for virtual reality providing the accuracy and scalability usually found in much more expensive tracking systems. PTrack is composed of single camera tracking PTrack Units. Each unit is connected to a video camera equipped with infrared strobes and features a novel iterative geometric pose estimation algorithm which does marker-based single camera tracking and is therefore completely autonomous. Multiple PTrack units successively extend the tracking range of the system. For a smooth transition of tracked labels from one camera to another, camera range areas must overlap to form a contiguous tracking space. A PTrack Sensor Fusion Module then computes the pose of a certain label within tracking space and forwards it to interested applications. A universal test setup for optical tracking systems has been built allowing to measure translational and rotational accuracy of PTrack as well as of competing systems.

Extensible middleware framework for multimodal interfaces in distributed environments

Abstract: We present a framework to manage multimodal applications and interfaces in a reusable and extensible manner. We achieve this by focusing the architecture both on applications' needs and devices' capabilities. One particular domain we want to approach is collaborative environments where several modalities and applications make it necessary to provide for an extensible system combining diverse components across heterogeneous platforms on-the-fly. This paper describes the proposed framework and its main contributions in the context of an architectural application scenario. We demonstrate how to connect different non-conventional applications and input modalities around an immersive environment (tiled display wall).

Follow me: Communicating intentions with a spherical robot

Abstract: In recent years, robots have gradually become incorporated in our society and therefore play more relevant role in social environments. These robots vary in form, some being more anthropomorphic than others. This, creates a need to study their interaction with the world. In this paper we used Sphero and BB-8, two robots with a simple spherical body devoid of verbal and other complex communication methods, to investigate how they can communicate intention to people. A set of behaviors based on pet behaviors was designed and tested in a controlled experiment, where the robot's aim was to convince a participant to follow it. We concluded that the use of these behaviors allows a robot to effectively communicate intention as well as create a bond with the participant, who would treat it as an equal, thereby engaging it in social interactions such as playing with it or talking to it.