Catálogo de publicaciones - libros

Many people believe that emotions and subjective feelings are one and the same and that a goal of human-centered computing is emotion recognition. The first belief is outdated; the second mistaken. For human-centered computing to succeed, a different way of thinking is needed. Emotions are species-typical patterns that evolved because of their value in addressing fundamental life tasks. Emotions consist of multiple components, of which subjective feelings may be one. They are not directly observable, but inferred from expressive behavior, self-report, physiological indicators, and context. I focus on expressive facial behavior because of its coherence with other indicators and research. Among the topics included are measurement, timing, individual differences, dyadic interaction, and inference. I propose that design and implementation of perceptual user interfaces may be better informed by considering the complexity of emotion, its various indicators, measurement, individual differences, dyadic interaction, and problems of inference.

Instinctive computing is a computational simulation of biological and cognitive instincts. It is a meta-program of life, just like universal gravity in nature. It profoundly influences how we look, feel, think, and act. If we want a computer to be genuinely intelligent and to interact naturally with us, we must give computers the ability to recognize, understand, even primitive instincts. In this paper, we will review the recent work in this area, the building blocks for the instinctive operating system, and potential applications. The paper proposes a ’bottom-up’ approach that is focused on human basic instincts: forage, vigilance, reproduction, intuition and learning. They are the machine codes in human operating systems, where high-level programs, such as social functions can override the low-level instinct. However, instinctive computing has been always a default operation. Instinctive computing is the foundation of Ambient Intelligence as well as Empathic Computing. It is an essential part of Human Computing.

A widely accepted prediction is that computing will move to the background, weaving itself into the fabric of our everyday living spaces and projecting the human user into the foreground. If this prediction is to come true, then next generation computing should be about anticipatory user interfaces that should be human-centered, built for humans based on human models. They should transcend the traditional keyboard and mouse to include natural, human-like interactive functions including understanding and emulating certain human behaviors such as affecti0ve and social signaling This article discusses how far are we from enabling computers to understand human behavior.

Automatic multimodal recognition of spontaneous emotional expressions is a largely unexplored and challenging problem. In this paper, we explore audio-visual emotion recognition in a realistic human conversation setting—the Adult Attachment Interview (AAI). Based on the assumption that facial expression and vocal expression are at the same coarse affective states, positive and negative emotion sequences are labeled according to Facial Action Coding System. Facial texture in visual channel and prosody in audio channel are integrated in the framework of Adaboost multi-stream hidden Markov model (AdaMHMM) in which the Adaboost learning scheme is used to build component HMM fusion. Our approach is evaluated in AAI spontaneous emotion recognition experiments.

Affective and human-centered computing have attracted a lot of attention during the past years, mainly due to the abundance of devices and environments able to exploit multimodal input from the part of the users and adapt their functionality to their preferences or individual habits. In the quest to receive feedback from the users in an unobtrusive manner, the combination of facial and hand gestures with prosody information allows us to infer the users’ emotional state, relying on the best performing modality in cases where one modality suffers from noise or bad sensing conditions. In this paper, we describe a multi-cue, dynamic approach to detect emotion in naturalistic video sequences. Contrary to strictly controlled recording conditions of audiovisual material, the proposed approach focuses on sequences taken from nearly real world situations. Recognition is performed via a ’Simple Recurrent Network’ which lends itself well to modeling dynamic events in both user’s facial expressions and speech. Moreover this approach differs from existing work in that it models user expressivity using a dimensional representation of activation and valence, instead of detecting discrete ’universal emotions’, which are scarce in everyday human-machine interaction. The algorithm is deployed on an audiovisual database which was recorded simulating human-human discourse and, therefore, contains less extreme expressivity and subtle variations of a number of emotion labels.

Computer models can be used to investigate the role of emotion in learning. Here we present , our framework for the systematic study of the relation between motion, daptation and einforcement earning (RL). EARL enables the study of, among other things, communicated affect as reinforcement to the robot; the focus of this chapter. In humans, emotions are crucial to learning. For example, a parent—observing a child—uses emotional expression to encourage or discourage specific behaviors. Emotional expression can therefore be a reinforcement signal to a child. We hypothesize that affective facial expressions facilitate robot learning, and compare a setting with a one to test this. The non-social setting consists of a simulated robot that learns to solve a typical RL task in a continuous grid-world environment. The social setting additionally consists of a human (parent) observing the simulated robot (child). The human’s emotional expressions are analyzed in real time and converted to an additional reinforcement signal used by the robot; positive expressions result in reward, negative expressions in punishment. We quantitatively show that the “social robot” indeed learns to solve its task significantly faster than its “non-social sibling”. We conclude that this presents strong evidence for the potential benefit of affective communication with humans in the reinforcement learning loop.

This work addresses the problem of human action recognition by introducing a representation of a human action as a collection of short trajectories that are extracted in areas of the scene with significant amount of visual activity. The trajectories are extracted by an auxiliary particle filtering tracking scheme that is initialized at points that are considered salient both in space and time. The spatiotemporal salient points are detected by measuring the variations in the information content of pixel neighborhoods in space and time. We implement an online background estimation algorithm in order to deal with inadequate localization of the salient points on the moving parts in the scene, and to improve the overall performance of the particle filter tracking scheme. We use a variant of the Longest Common Subsequence algorithm (LCSS) in order to compare different sets of trajectories corresponding to different actions. We use Relevance Vector Machines (RVM) in order to address the classification problem. We propose new kernels for use by the RVM, which are specifically tailored to the proposed representation of short trajectories. The basis of these kernels is the modified LCSS distance of the previous step. We present results on real image sequences from a small database depicting people performing 12 aerobic exercises.

This chapter addresses the problem of multimodal analysis of human face–to–face communication. This is imporant since in the near future, smart environments equipped with multiple sensory systems will be able to sense the presence of humans and assess recognize their behaviours, actions, and emotional states. The main goal of the presented study is to develop models of communicative/interactive events in multimedia (audio and video), suitable for the analysis and subsequent incorporation within virtual reality environments. Interactive, environmental, and emotional characteristics of the communicators are estimated in order to define the communication event as one entity. This is achieved by putting together results obtained in social sciences and multimedia signal processing under one umbrella – the communication atmosphere analysis. Experiments based on real life recordings support the approach.

A common problem of ubiquitous sensor-network computing is combining evidence between multiple agents or experts. We demonstrate that the , our novel formulation for combining evidence from multiple dynamic classification processes (“experts”), can achieve greater accuracy, efficiency, and robustness to data corruption than standard methods such as HMMs. It accomplishes this by simultaneously modeling the structure of interaction and the latent states.

The central concern of Social Intelligence Design is the under-standing and augmentation of social intelligence that might be attributed to both an individual and a group. Social Intelligence Design addresses understanding and augmentation of social intelligence resulting from bilateral interaction of intelligence attributed to an individual to coordinate her/his behavior with others in a society and that attributed to a collection of individuals to achieve goals as a whole and learn from experiences. Social intelligence can be addressed from multiple perspectives. In this chapter, I will focus on three aspects. First, I highlight interaction from the social discourse perspective in which social intelligence manifests in rapid interaction in a small group. Second, I look at the community media and social interaction in the large, where slow and massive interaction takes place in a large collection of people. Third, I survey work on social artifacts that embody social intelligence. Finally, I attempt to provide a structured view of the field.