Augmented attribute representations

We propose a new learning method to infer a mid-level feature representation that combines the advantage of semantic attribute representations with the higher expressive power of non-semantic features. The idea lies in augmenting an existing attribute-based representation with additional dimensions for which an autoencoder model is coupled with a large-margin principle. This construction allows a smooth transition between the zero-shot regime with no training example, the unsupervised regime with training examples but without class labels, and the supervised regime with training examples and with class labels. The resulting optimization problem can be solved efficiently, because several of the necessity steps have closed-form solutions. Through extensive experiments we show that the augmented representation achieves better results in terms of object categorization accuracy than the semantic representation alone. © 2012 Springer-Verlag.

Encoding of mixtures in a simple olfactory system.

Natural odors are usually mixtures; yet, humans and animals can experience them as unitary percepts. Olfaction also enables stimulus categorization and generalization. We studied how these computations are performed with the responses of 168 locust antennal lobe projection neurons (PNs) to varying mixtures of two monomolecular odors, and of 174 PNs and 209 mushroom body Kenyon cells (KCs) to mixtures of up to eight monomolecular odors. Single-PN responses showed strong hypoadditivity and population trajectories clustered by odor concentration and mixture similarity. KC responses were much sparser on average than those of PNs and often signaled the presence of single components in mixtures. Linear classifiers could read out the responses of both populations in single time bins to perform odor identification, categorization, and generalization. Our results suggest that odor representations in the mushroom body may result from competing optimization constraints to facilitate memorization (sparseness) while enabling identification, classification, and generalization.

Encoding of mixtures in a simple olfactory system

Natural odors are usually mixtures; yet, humans and animals can experience them as unitary percepts. Olfaction also enables stimulus categorization and generalization. We studied how these computations are performed with the responses of 168 locust antennal lobe projection neurons (PNs) to varying mixtures of two monomolecular odors, and of 174 PNs and 209 mushroom body Kenyon cells (KCs) to mixtures of up to eight monomolecular odors. Single-PN responses showed strong hypoadditivity and population trajectories clustered by odor concentration and mixture similarity. KC responses were much sparser on average than those of PNs and often signaled the presence of single components in mixtures. Linear classifiers could read out the responses of both populations in single time bins to perform odor identification, categorization, and generalization. Our results suggest that odor representations in the mushroom body may result from competing optimization constraints to facilitate memorization (sparseness) while enabling identification, classification, and generalization

Sensing through body dynamics

It has been shown that sensory morphology and sensory-motor coordination enhance the capabilities of sensing in robotic systems. The tasks of categorization and category learning, for example, can be significantly simplified by exploiting the morphological constraints, sensory-motor couplings and the interaction with the environment. This paper argues that, in the context of sensory-motor control, it is essential to consider body dynamics derived from morphological properties and the interaction with the environment in order to gain additional insight into the underlying mechanisms of sensory-motor coordination, and more generally the nature of perception. A locomotion model of a four-legged robot is used for the case studies in both simulation and real world. The locomotion model demonstrates how attractor states derived from body dynamics influence the sensory information, which can then be used for the recognition of stable behavioral patterns and of physical properties in the environment. A comprehensive analysis of behavior and sensory information leads to a deeper understanding of the underlying mechanisms by which body dynamics can be exploited for category learning of autonomous robotic systems. © 2006 Elsevier Ltd. All rights reserved.

语音意识缺陷儿童的言语知觉技能

Considerable studies find that developmental dyslexia is associated with deficits in phonological processing skills, especially phonological awareness. In order to explore the nature of phonological awareness deficits in dyslexia, researchers have begun to investigate the role of speech perception. The findings about speech perception abilities in dyslexics are inconsistent. The heterogeneity of dyslexia may be responsible for the inconsistency of findings. Considering the general suggestion that phonological awareness deficits in dyslexia are attributed to categorical perception deficits, it is more direct to examine whether children with phonological awareness difficulties or phonological dyslexia show speech categorization deficits consistently. The present study would investigate whether Chinese children with phonological awareness deficits or phonological dyslexia showed abnormal speech perception. The whole study consisted of two parts. Part I screened children with phonological-awareness deficits from Year 3 kindergartens and examined their abilities of perceiving native category continuum, nonnative category contrasts and non-speech sound series. Part II selected phonological dyslexics from an elementary school as participants, and further explored the relation between phonological deficits and speech perception. The first two experiments of Part II examined separately the abilities to label stimuli in native category continuum and brief stops in different contexts, the last experiment investigated the adaptation effects of different participant groups. The main conclusions are as follows: 1) Children with phonological dyslexia showed categorical perception deficits: they had lower consistency than controls when perceiving stimuli within phonetic categories, especially for the stimuli which were not natural sounds. 2) Children with phonological dyslexia exhibited a general difficulty of perceiving brief segments of stops from different contexts. 3) Children with phonological dyslexia did not show adaptation to repeatedly presented stimuli. Based on the present conclusions and the findings of previous studies, we suggested that the representations of sound stimuli in phonological dyslexics’ brains are different from those in normal children’s; the representations of sound stimuli in dyslexics’ cortical neural networks are more diffuse and inconsistent.

社会困境中的行为决策――社会身份与地位的影响

Social dilemmas are defined as such situations in which short-term individual and long-term collective interests are at odds (Hardin, 1968; Messick & Brewer, 1983). Severe social problems lead more and more researchers to pay attention to the issue of social dilemmas. Until now, research has widely examined the factors influencing cooperation in social dilemmas, and provided solutions to social dilemmas. In the need of theory development, previous research generally simplified the situation of realistic social dilemmas. Therefore, few studies have explored the issue of cooperation in social dilemmas in an intergroup situation. On the other hand, there has been little empirical attention for such issues in countries outside of United States and Western Europe. To meet this gap, the present study grounded itself in social identity theories and examined the effect of unequal group status on behavioral decision making in social dilemmas. To this end, the study designed three experiments to examine how individuals with different group status response psychologically and behaviorally to cooperation in intragroup and intergroup social dilemmas. In experiment 1, the study examined how independent and interdependent self- construals affect cooperative behavior in an intragroup social dilemma. The results showed that when individuals were primed with interdependent (as opposed to independent) self-construal, they consistently contributed highly, regardless of context manipulation. In contrast, those primed with independent self-construal contributed less in the investment game but only when placed in a context where group members were encouraged to think about their individual (versus shared) fate. Results supported the idea that independent self-construal in a low interdependent context produces the most competitive behavior and that this effect was partially mediated by the feeling of interaction within a group. In experiment 2, the study examined how the effect of group status on different level of cooperation in a nested social dilemma was moderated by individual status, and what roles ingroup and superordinate identifications played in the above effect. Results found that individuals in higher status groups tended to allocate more to private account and less to subgroup account compared to those in lower status group; individuals in higher status groups allocated more to higher level accounts than to private account, whereas those in lower status acted in a reverse way. The results indicated that group status (compared to individual status) exerted a positive influence on behavioral decision making in social dilemmas, with higher group status contributing more to subgroup as well as collective interests. Results also found that the effect of group status and/or individual status on cooperation in social dilemmas was moderated by sex. As for individual status, results showed that the effect of individual status on subgroup interest was significantly moderated only by the combination of higher ingroup identification and lower superordinate identification. In experiment 3, the study explored how group stability and cognitive categorization interactively influenced the relationship between group status and behavioral decision making in a social dilemma. Results did not support the prediction that group status and stability interactively affected behavioral decision-making in social dilemmas. However, it was found that this relationship was moderated by which level individuals categorized themselves at. When categorization at the individual level was salient, individuals in high status group contributed more to subgroup account than those in low status group if they perceived a stable status hierarchy; whereas they contributed more to private account and less to collective account if they perceived that the status was instable. On the other hand, when categorization at the subgroup level was salient, individuals in high status group contributed less to collective account than those in low status group if they perceived that the status was stable; whereas they contributed less to private account and more to subgroup account if they perceived an instable status relation. In summary, the present study suggests that cooperation with ingroup forms the basis of social harmony, and higher status for everyone in any given group is a necessary for social development. On the other hand, individuals in higher status group tend to be more selfish once they realize that their current status hierarchy is unstable. However, activating their collective identity will to some degree increase the level of their cooperation with the collective. The study thus provides psychological explanations on how to construct group harmony and management suggestions on how to solve social conflicts.

空间关系类别的形成与计算

Spatial categorization and computation were studied systematically and sequentially in the dissertation, in which, category learning paradigm was transferred to spatial domain. Four parts were developed in our study, in which, the former two parts investigated the mechanism of spatial categorization, and the latter two parts explored two applied issues. Spatial category formation was investigated by two aspects in distance and angle respectively in research one. The results showed that subjects can make accurate judgment and rating for corresponding spatial categorical prototypes and boundaries. In the second research, the formation of spatial relationship categorization was examined through two factors of number of categories and different spatial configurations. The results were repeated in this part, and also, different influences of reference of frame were found in the process of spatial categorization. In the third research, the influences of spatial categorization on spatial localization were found, that in subject could make different location judgment with new acquired spatial categories. Spatial position distribution has no effect on spatial localization on the other hand. The result that reference objects have significant influences on applicability of spatial relationship was found in research four. Our Bayesian hierarchical model could be applied in simulation and prediction of perception of “above” spatial relationship. Theoretical significance and applied importance were discussed in detail.

情绪对创造力影响的发展性特点

Although robust findings have been that positive mood leads to greater creativity, several other literatures have found that negative mood sometimes results in more creative performance than positive or neutral mood. Several possible explanations for this emotion–creativity link have been proposed by researchers, but there has not yet been definitive research identifying the mechanism(s) behind this relationship. This research represents an initial step in this direction, examining the possibility that Intelligence Current may be a contributing mechanism in the emotion–creativity link from the perspective of development. The object of the present study was to do the followings 1) the effects and mechanism(s) of emotions on creativity development from adolescence to young adulthood by Unusual Uses Test; 2) the possibility that Intelligence Current may be a contributing mechanism in the emotion–creativity link. The participants were 849 adolescents in high schools and 267 undergraduates in the university aged from 11 to 22 years old. The mechanism(s) for emotion-creativity link was explored by cognitive flexibility (assessed using Abstract Match Task), tolerance (inclusive) ratings (assessed using Categorization of Analogy Task), uses originality ratings, and confidence ratings. Results indicated that: 1) The level of creativity varied with age. It increased from 11 years on, but decreased at about 14 to improve again from 15 to 22 years. 2) The different effects of emotions on creativity development among adolescents and undergraduates emerged, but the effects of positive and negative emotions on creativity didn’t differ from each other for the whole participants. Furthermore, compared with positive and negative emotions, the neutral emotions produced the lowest creativity for 11.00-13.99 years old group, but produced the highest creativity for 14.00-14.99 and 17.00-21.99years old. 3) Positive emotions have been shown to enable individuals to higher level of cognitive flexibility and better performance than negative emotions by Abstract Match Task, which could be considered that the Intelligence Current may be a contributing mechanism in the emotion–creativity link. 4) Positive emotions have been shown to enable individuals to own higher confidence, to categorize items with more flexibility, to see more potential relatedness among unusual and atypical members of categories, to evaluate items more originality than negative emotions among most of participants, especially the group of 14.00-14.99 years old. In sum, the present study helps us to further understand that the term ‘Intelligence Current’ is further explained and the problems found in relationships between creativity and emotions. However, further research is required to explore and confirm the conclusions of the present study.

颜色在自然场景知觉中的作用

Color has an unresolved role in the rapid process of natural scene. The temporal changes of the color effect might partly account for the debates. Besides, the distinction of localized and unlocalized information has not been addressed directly in these color studies. Here we present two experiments that investigate whether color contributes to categorization in a briefly flashed natural image and also whether it is mediated by time and low-level information. By controlling the interval between target and mask stimuli, Experiment 1 tested the hypothesis that colors could facilitate in the early stage of scene perception and the effect would decay in later processing. Experiment 2 examined how the randomization of local phase information influenced the color’s advantage over gray. Together, the results suggest that color does enhance natural scene categorization at short exposure time. Furthermore, results imply that effect of color is stable between 12 and120ms, and is not accounted by showing the structures organized by localized information. Therefore，we concluded that color always make effect in the process of rapid scene categorization, and do not depend on localized information. Thus, the present study is an attempt to fill the gap in previous research; its results is an contribution to deeper understanding of the role of color in natural scene perception.

中美大学生归因偏差的跨文化研究

This research systematically compared Chinese undergraduates with American undergraduates on four kinds of attributional bias: correspondent bias, siuational overattribution, intergroup attributional bias and self-serving attributional bias, and examined the effect of the implicit theories reflecting the cultures on attributional bias. First is analyzed three pairs of opposite implicit theories: dispositionalist theory and situationalist theory, generality and particularity, stressing the positive evaluation of self and despising the positive evaluation of self. It developed the Modern Implicit Theories Inventory and Traditional Implicit Theories Inventory to measure these implicit theories, and the results showed these inventories had satisfactory validity and reliability, and they were suitable for the group comparison of Chinese implicit theories with European-American. At the same time through the test it found Chinese undergraduates agreed all these opposite implicit theories more than American undergraduates. Second, it studied Chinese and American undergraduates' attributional accuracy on locus of causality. The results showed: Chinese and American undergraduates both had the correspondent bias under the different salient situational constraints, and the degree of Chinese and American undergraduates' correspondent bias under the different salient situational constraints had no significant difference' Chinese and American undergraduates both showed the situational overattribution; Chinese undergraduates had more the correspondent bias and situational overattribution than American undergraduates. Third, on the research of Chinese and American undergraduates' intergroup attributional bias, it found Chinese and American undergraduates both had no intergroup attributional bias among kin, friends and strangers, while they both show some favorable outcome effects for these three group actors. The favorable outcome effects were significant on the attributional dimensions of locus of causality and controllability for strangers' behavior, and stability for kin and friends' behavior rating by Chinese undergraduates, and stability for friends' behavior rating by American undergraduates. Fourth, it explored Chinese and American undergraduates' self-serving attributional bias, and the result indicated that Chinese and American undergraduates both showed significant self-serving attributional bias: for outcome effects, Chinese undergraduates' self-serving attributional bias were reflected on the attributional dimensions of locus of causality, stability, controllability and globality, and American undergraduates were reflected on the attributional dimensions of locus of causality, stability and globality; for categorization effects, both Chinese and American undergraduates' self-serving attributional bias were reflected on attributional difference between self's negative behavior and others', but Chinese undergraduates were embodied on the attributional dimensions of locus of causality, stability and globality while American undergraduates were reflected on the attributional dimensions of stability and globality. It also found Chinese undergraduates had more self-serving attributional bias than American undergraduates. This was reflected on the attributional dimensions of locus of causality, stability and controllability for outcome effects, and for categorization effects, locus of causality, stability and globality rating for self and others' negative behavior. All studies indicated that Chinese and American undergraduates' implicit theories had no significant effects on all their four attributional bias. These findings' potentially important implications were discussed and the further research was suggested.

Recognition by Prototypes

A scheme for recognizing 3D objects from single 2D images is introduced. The scheme proceeds in two stages. In the first stage, the categorization stage, the image is compared to prototype objects. For each prototype, the view that most resembles the image is recovered, and, if the view is found to be similar to the image, the class identity of the object is determined. In the second stage, the identification stage, the observed object is compared to the individual models of its class, where classes are expected to contain objects with relatively similar shapes. For each model, a view that matches the image is sought. If such a view is found, the object's specific identity is determined. The advantage of categorizing the object before it is identified is twofold. First, the image is compared to a smaller number of models, since only models that belong to the object's class need to be considered. Second, the cost of comparing the image to each model in a classis very low, because correspondence is computed once for the whoel class. More specifically, the correspondence and object pose computed in the categorization stage to align the prototype with the image are reused in the identification stage to align the individual models with the image. As a result, identification is reduced to a series fo simple template comparisons. The paper concludes with an algorithm for constructing optimal prototypes for classes of objects.

Context-Based Vision System for Place and Object Recognition

While navigating in an environment, a vision system has to be able to recognize where it is and what the main objects in the scene are. In this paper we present a context-based vision system for place and object recognition. The goal is to identify familiar locations (e.g., office 610, conference room 941, Main Street), to categorize new environments (office, corridor, street) and to use that information to provide contextual priors for object recognition (e.g., table, chair, car, computer). We present a low-dimensional global image representation that provides relevant information for place recognition and categorization, and how such contextual information introduces strong priors that simplify object recognition. We have trained the system to recognize over 60 locations (indoors and outdoors) and to suggest the presence and locations of more than 20 different object types. The algorithm has been integrated into a mobile system that provides real-time feedback to the user.

Webpage Classification with ACO-enhanced Fuzzy-Rough Feature Selection.

R. Jensen and Q. Shen, 'Webpage Classification with ACO-enhanced Fuzzy-Rough Feature Selection,' Proceedings of the Fifth International Conference on Rough Sets and Current Trends in Computing (RSCTC 2006), LNAI 4259, pp. 147-156, 2006.

Modal Matching for Correspondence and Recognition

Modal matching is a new method for establishing correspondences and computing canonical descriptions. The method is based on the idea of describing objects in terms of generalized symmetries, as defined by each object's eigenmodes. The resulting modal description is used for object recognition and categorization, where shape similarities are expressed as the amounts of modal deformation energy needed to align the two objects. In general, modes provide a global-to-local ordering of shape deformation and thus allow for selecting which types of deformations are used in object alignment and comparison. In contrast to previous techniques, which required correspondence to be computed with an initial or prototype shape, modal matching utilizes a new type of finite element formulation that allows for an object's eigenmodes to be computed directly from available image information. This improved formulation provides greater generality and accuracy, and is applicable to data of any dimensionality. Correspondence results with 2-D contour and point feature data are shown, and recognition experiments with 2-D images of hand tools and airplanes are described.