Emotion and Adaptation

Modeling visual attention-particularly stimulus-driven, saliency-based attention-has been a very active research area over the past 25 years. Many different models of attention are now available which, aside from lending theoretical contributions to other fields, have demonstrated successful applications in computer vision, mobile robotics, and cognitive systems. Here we review, from a computational perspective, the basic concepts of attention implemented in these models. We present a taxonomy of nearly 65 models, which provides a critical comparison of approaches, their capabilities, and shortcomings. In particular, 13 criteria derived from behavioral and computational studies are formulated for qualitative comparison of attention models. Furthermore, we address several challenging issues with models, including biological plausibility of the computations, correlation with eye movement datasets, bottom-up and top-down dissociation, and constructing meaningful performance measures. Finally, we highlight current research trends in attention modeling and provide insights for future.

State-of-the-Art in Visual Attention Modeling

Regression-based statistical mediation and moderation analysis in clinical research: Observations, recommendations, and implementation.

http://n-acetylaspartate.com/PDF%20files/NAA%20Review.pdf

N-Acetylaspartate in the CNS: From neurodiagnostics to neurobiology

When natural scenes are viewed, a multitude of objects that are stable in their environments are brought in and out of view by eye movements. The posterior parietal cortex is crucial for the analysis of space, visual attention and movement 1 . Neurons in one of its subdivisions, the lateral intraparietal area (LIP), have visual responses to stimuli appearing abruptly at particular retinal locations (their receptive fields)2. We have tested the responses of LIP neurons to stimuli that entered their receptive field by saccades. Neurons had little or no response to stimuli brought into their receptive field by saccades, unless the stimuli were behaviourally significant. We established behavioural significance in two ways: either by making a stable stimulus task-relevant, or by taking advantage of the attentional attraction of an abruptly appearing stimulus. Our results show that under ordinary circumstances the entire visual world is only weakly represented in LIP. The visual representation in LIP is sparse, with only the most salient or behaviourally relevant objects being strongly represented.

The representation of visual salience in monkey parietal cortex

Decreased Spontaneous Attention to Social Scenes in 6-Month-Old Infants Later Diagnosed with Autism Spectrum Disorders

In this study we examined the social behaviors of 4- to 12-year-old children with autism spectrum disorders (ASD; N = 24) during three tradic interactions with an adult confederate and an interaction partner, where the interaction partner varied randomly among (1) another adult human, (2) a touchscreen computer game, and (3) a social dinosaur robot. Children spoke more in general, and directed more speech to the adult confederate, when the interaction partner was a robot, as compared to a human or computer game interaction partner. Children spoke as much to the robot as to the adult interaction partner. This study provides the largest demonstration of social human-robot interaction in children with autism to date. Our findings suggest that social robots may be developed into useful tools for social skills and communication therapies, specifically by embedding social interaction into intrinsic reinforcers and motivators.

/pdf/social-robots-as-embedded-reinforcers-of-social-behavior-in-1foyw3synl.pdf

Social robots as embedded reinforcers of social behavior in children with autism.

Young children with autism spend relatively little time monitoring other people and the facial expressions and gestures of others (Dawson et al., 2004; Swettenham et al., 1998). Given that the acquisition of language and the development of social cognition are highly experience-dependent processes (Greenough, Black, & Wallace, 1987), deficits in the ability to select for processing and to sustain attention on faces early in development are likely to have a detrimental effect on early socio-cognitive (Leppanen & Nelson, 2009; Pascalis et al., 2005) and language development (Kuhl, Williams, Lacerda, Stevens, & Lindblum,1992; Thiessen, Hill, & Saffran, 2007). Furthermore, given the role observational learning plays in the development of nonsocial cognition, atypical attention towards people is likely to result in collateral difficulties in other domains as well (Shic, Bradshaw, Klin, Scassellati, & Chawarska, 2011). Thus, it is essential to advance our understanding of the processes that underlie atypical social attention at the earliest ages when autism can be reliably identified. The benefit of such an approach is two-fold: it will advance identification of target skills for intervention, and will inform the design of methods for screening infants at risk for autism spectrum disorders (ASDs) in the first year of life.

In the past decade, studies utilizing eye movements as indices of perception and attention have shown great promise for identifying mechanisms underlying social disability in autism. A majority of studies have employed static images of faces and people, sometimes presented in isolation from their social context. This line of research has generated important insights into the attentional, perceptual, and learning strategies associated with autism in the first three years of life. Empirical evidence suggests that faces do not capture the attention of toddlers with ASD as readily as they capture the attention of non-affected toddlers (Coffman, Shic, Meltvedt, Bradshaw & Chawarska, 2011). When examining a novel face, toddlers with ASD exhibit atypical scanning patterns (Chawarska & Shic, 2009) and require more time to extract invariant face features necessary for recognition (Bradshaw, Shic, & Chawarska, 2011; Chawarska & Shic, 2009; Chawarska & Volkmar, 2007; Webb et al. 2010).

Although studies examining the processing of static faces are essential for parsing the nature of deficits in autism and reflect upon the atypical functioning of cortical networks involved in the structural analysis of faces (Haxby, Hoffman, & Gobbini, 2002), in real life faces present as dynamic stimuli. This aspect of face processing relies heavily upon multisensory brain areas involved in the perception of biological motion, speech, and social cognition such as the posterior superior temporal sulcus (pSTS) as well as the attentional network involved in selection of and sustained attention to these salient stimuli (Allison, Puce, & McCarthy, 2000; Calvert & Campbell, 2003; Haxby et al., 2002). Several eye-tracking studies have examined the attentional responses to dynamic faces embedded in complex naturalistic contexts in toddlers with ASD (Jones, Carr, & Klin, 2008; Shic, et al., 2011). These studies suggest that when toddlers with ASD view videos of adults trying to engage them through simple social games (e.g., peek-a-boo) (Jones et al., 2008) or observe parent-child dyads engaged in a shared activity (Shic et al., 2011), they tend to look at faces less than control participants, a pattern similar to that seen in high-functioning adolescents with ASD (Klin, Jones, Schultz, Volkmar, & Cohen, 2002). The reasons for limited face monitoring in dynamic displays are not clear. It is possible that toddlers with autism spend less time attending to faces as means of regulating their arousal (Hutt & Ounsted,1966). It is also possible that they spend less time monitoring faces simply because faces are not prioritized in their attentional system for processing to the same extent as faces are in non-autistic individuals (Chawarska, Klin, & Volkmar,2003; Chawarska, Volkmar, & Klin, 2010). Finally, it is possible that limited attention to people is a direct result of increased salience of objects, whether it be due to high-level (e.g., cars) (Sasson, Turner-Brown, Holtzclaw, Lam, & Bodfish, 2008) or low-level (e.g., high contrast) attributes (Bertone, Mottron, Jelenic, & Faubert, 2005; McCleery, Allman, Carver, & Dobkins, 2007; Shic, Chawarska, Lin, & Scassellati, 2007).

In the present study we examine the effect of context on the regulation of attention in toddlers with autism (AUT) as well as developmentally delayed (DD) and typically developing (TD) controls using dynamic social scenes. To parse the factors that are responsible for limited attention to faces we manipulated context with regard to the presence of salient social (child-directed speech and eye contact) and nonsocial (distractor toys) cues. If toddlers with autism exhibit a generalized social attention deficit, they would spend, compared to non-ASD controls, less time examining the person’s face in all contexts; instead we would expect their attention to be directed toward toys in the background. However, if they show elementary sensitivity to the context in which people appear in their visual field, then the atypical patterns should be condition-specific. In this case we would expect the smallest differences between groups in the condition with highly salient non-social distractors (moving mechanical toys) and the greatest differences in the condition containing explicit and prolonged social bids, including eye contact and child-directed speech.

Context modulates attention to social scenes in toddlers with autism

This study used eye-tracking to examine visual scanning and recognition of faces by 2- and 4-year-old children with autism spectrum disorder (ASD) (N = 44) and typically developing (TD) controls (N = 30). TD toddlers at both age levels scanned and recognized faces similarly. Toddlers with ASD looked increasingly away from faces with age, atypically attended to key features of faces, and were impaired in face recognition. Deficits in recognition were associated with imbalanced attention between key facial features. This study illustrates that face processing in ASD may be affected early and become further compromised with age. We propose that deficits in face processing likely impact the effectiveness of toddlers with ASD as social partners and thus should be targeted for intervention.

/pdf/looking-but-not-seeing-atypical-visual-scanning-and-3w5iis8trx.pdf

Looking But Not Seeing: Atypical Visual Scanning and Recognition of Faces in 2 and 4-Year-Old Children with Autism Spectrum Disorder

Social robots can offer tremendous possibilities for autism spectrum disorder (ASD) interventions. To date, most studies with this population have used short, isolated encounters in controlled laboratory settings. Our study focused on a 1-month, home-based intervention for increasing social communication skills of 12 children with ASD between 6 and 12 years old using an autonomous social robot. The children engaged in a triadic interaction with a caregiver and the robot for 30 min every day to complete activities on emotional storytelling, perspective-taking, and sequencing. The robot encouraged engagement, adapted the difficulty of the activities to the child’s past performance, and modeled positive social skills. The system maintained engagement over the 1-month deployment, and children showed improvement on joint attention skills with adults when not in the presence of the robot. These results were also consistent with caregiver questionnaires. Caregivers reported less prompting over time and overall increased communication.

Frederick Shic

Papers

Decreased Spontaneous Attention to Social Scenes in 6-Month-Old Infants Later Diagnosed with Autism Spectrum Disorders

Social robots as embedded reinforcers of social behavior in children with autism.

Context modulates attention to social scenes in toddlers with autism

Looking But Not Seeing: Atypical Visual Scanning and Recognition of Faces in 2 and 4-Year-Old Children with Autism Spectrum Disorder

Improving social skills in children with ASD using a long-term, in-home social robot