Figure 2 - uploaded by Tony Belpaeme
Content may be subject to copyright.
1: The visual pathways, (1) the optic nerve, (2) the optic chiasm, (3) optic tract, (4) lateral geniculate nucleus and (5) the visual cortex. Adapted from (Cotter, 1990).
Source publication
van de academische graad van doctor in de wetenschappen, in het openbaar te verdedigen op vrijdag 8 maart 2002. Acknowledgements I started as a research assistant in the Artificial Intelligence Laboratory in autumn 1996. My first interests were into behavioural robotics and robot ecosystems. As a continuation to my “licentiaats ” thesis I started b...
Contexts in source publication
Context 1
... signals are then conveyed along the optic tract to the visual cortex. Figure 2.1 shows a horizontal cross-section of the brain with the visual pathway. Light falls on the retina in the eye, and initiates a signal travelling down the optic nerve. ...
Context 2
... the visual cor- tex we find more parallel visual streams, but they are not of concern for this work; detailed accounts can be found in for example (Cornsweet, 1970;Cotter, 1990;Leibovic, 1990;Zeki, 1993;Arbib, 1995;Kaiser and Boynton, 1996). Figure 2.2 shows a schematic cross section of the human eye. ...
Context 3
... and cone recep- tors serve very different functions. Rods are involved in low light vision (called scotopic vision), their reaction time is slow, but their sensitivity is Figure 2.3: Rod and cone density as a function of the retinal location; clearly show- ing how there are hardly any rod photoreceptors in the fovea, while the cone den- sity is very high. ...
Context 4
... roughly corresponds to red, green and blue. Figure 2.4 shows the relative absorbance for the four human photoreceptor pigments. ...
Context 5
... opponent channels 4 . So, opponent-colour processing is important for decorrelating the cone signals: as the responses of the cone cells are not sufficient for seeing colour, Figure 2.5: Encoding the cone signals into opponent-process signals. ...
Context 6
... best examples clustered in discrete regions on the colour chart, meaning that foci of basic colour terms agreed well for each language. Figure 2.7 shows the foci of the BCTs of the twenty languages reported. The concurrence is striking: not only do the foci lie in clustered regions; the areas enclosing the foci also don't overlap (only in the case of GREEN and BLUE there are BCTs which seem to encode for bluish green). ...
Context 7
... 7.25 illustrates the categori- cal repertoire of two agents at generation 200 of a population of 20 agents from the experiment in 7.3.2. There a strong resemblance between the two repertoires (which was already noticeable from the low category variance in figure 7.20), the two agents have had a common ancestor and only dif- fer because their genetic information has undergone some mutations since they both split off. ...
Similar publications
This paper presents our recent study on low-resource automatic speech recognition (ASR) system with accented speech. We propose multi-accent Subspace Gaussian Mixture Models (SGMM) and accent-specific Deep Neural Networks (DNN) for improving non-native ASR performance. In the SGMM framework, we present an original language weighting strategy to mer...
Citations
... Há, contemporaneamente, uma grande discussão sobre a natureza da percepção categórica, se é estritamente inata (SAUTER, et. al, 2011) ou pode ser aprendida (BELPAEME, 2002;NOTMAN, et. al, 2005). ...
Este artigo tem como objetivo mostrar que a discussão entre Universalismo das Cores e Relativismo Linguístico é apenas um debate linguístico e não podemos implicar através das teses que a percepção de cores é influenciada (ou não) pelo número de categorias para cores que uma língua dispõe. Para isso, analisamos minuciosamente as teses universalistas e relativistas a fim de encontrar evidências experimentais que suportem as implicações feitas: a língua influencia ou não a percepção de cores. Nenhum experimento proposto por linguistas suportam as implicações que fazem acerca da cognição e percepção humana. Teorias linguísticas não são suficientes para responder a pergunta acerca de como percebemos as cores e a relação entre língua e percepção.
... In previous research, preferences and meanings of colours have been suggested to vary according to cultural factors of religion (P1), nationality or nation state (P3), ethnicity (P4), and language (P5), (Belpaeme, 2002;Jacobs, Keown, Worthley & Ghymn, 1991;Lee, 1998;Madden, Hewett & Roth, 2000). According to the present empirical analyses, some of the most persisting colour meanings are, in fact, related to these four cultural forces, but also related to ideology/values (P2), and the environment (P6), which have less frequently been discussed in previous research. ...
In different cultures, colours are sometimes assigned different meanings. Understanding the origins of these cultural colour meanings has become increasingly important with the ongoing advances towards digitalization of business and communication, and the most recent phase of
globalization. However, academic research has largely neglected this phenomenon.
The purpose of this dissertation is to increase our understanding of the dynamics of crosscultural meaning-making for colours by examining the provenance of cultural colour meanings. This is achieved by employing empirical studies set in Chinese-Finnish business contexts, taking into account particular perspectives induced by globalization and online media, and their implications on the developments within the intersections of the research domains of business communication, culture, and colour.
Methodologically, this research applies an emic-etic cross-cultural approach, using within method triangulation of qualitative methods to explore the phenomenon. The empirical materials consist of narrative, visual, and observation materials generated within studies
conducted in China and in Finland between January 2007 and July 2009. The two countries were chosen due to the shift of interest of many Finland-based MNCs from China-as-a-factory to include China-as-a-market, and ensuing perception of previously unexplored cultural
differences in visual conventions, colour meanings in particular.
Engaging in the debate within management and communication studies this research suggests a potential synthesis between the systemic and the process views to Culture. It finds that the significance of some cultural forces is inherited and thus more persistent (systemic view),
while that of other cultural forces is more dynamic and transforming in nature (process view). Consequently, by framing Culture as dynamic heritage, this research proposes a conceptualization for Provenance of cultural colour meanings, contributing to the Multimodal theory of colour. Likewise, this research propositions the significance of the expressive
function of the mode of colour borrowing from the former art historical and psychoanalytic approaches as well as based on the present empirical studies. Consequently, engaging with the Interactionist theory of communication, this research suggests that Meaning-making for
colour can be conceived as an expression of cultural experience whereby the communicators become expressors of their respective cultural background and experience. Finally, this dissertation advances our understanding of meaning-making for colour in a multimodal
context by adding the cross-cultural dimension to previous models, and shows that different modes can elaborate, extend, enhance, and contradict each other.
... In a next step, it was shown how these ideas can be grounded in actual robotic agents (Steels 2001b;2002a;Steels & Vogt 1997;Vogt 2000). The structural coupling of concepts and lexicons has also been successfully applied to the domains of colour (Belpaeme 2002;Belpaeme & Bleys 2005;Steels & Belpaeme 2005) and space (Loetzsch et al. 2008b;. ...
... For constructing the hierarchical tree structure, non-metric distance functions for measuring the inter-class separability was developed in Lu et al. (2007Lu et al. ( , 2010 and Eiter and Mannila (1997). The significance of non-metric distance function in image classification and computer vision has been investigated in Jacobs, Weinshall, and Gdalyahu (2000), and the raison d'etre of non-metric distance function is also corroborated by some research in psychology suggesting the ubiquity of non-metric distance in human similarity judgements (Belpaeme, 2002;Niiniluoto, 1987). ...
... However, it is not weighted by the number of elements in each class. Enlightened by the weighted sum of minimum distances developed in Belpaeme (2002), the sum of minimum distances function was modified as d md ðA; BÞ ¼ ...
The synthesis of an effective multi-category nonlinear classifier with the capability to output calibrated posterior probabilities to enable post-processing is of great significance in practical recognition situations because the posterior probability reflects the assessment uncertainty. However, the estimation of posterior probability for multi-category classifiers is an unwieldy problem in the realm of pattern recognition, which usually is more intractable than that in dichotomic cases. In this paper, with the aid of binary tree representation for nested structures, a new polychotomous classification and posterior probability estimation scheme is developed on the strength of Bayesian decision theory. In particular, by capitalising on the intrinsic conexus between hierarchical structure and multi-scale analysis, the polychotomous multi-scale Bayesian kernel Fisher discriminant (KFD) is implemented for building the classifier at different scales for different levels. Finally, the performance of the proposed classification and posterior probability estimation algorithm is validated by designing a multi-category Bayesian KFD classifier for a benchmark satellite images dataset.
... In previous research (e.g. Belpaeme 2002), the weight of the interpretation variance of each form was equal to the number of agents (n) that know this form. However, this weight does not reflect the actual probability of such a form being used in a random interaction. ...
... In previous research, meanings of colors have been suggested to vary according to cultural factors such as religion, nationality, ethnicity, and language (Belpaeme 2002, Jacobs, Keown, Worthley & Ghymn 1991, Madden, Hewett & Roth 2000. Meanwhile, ideology, religion, and language have been found to shape cultural clusters (e.g. ...
This paper proposes a conceptual framework for analysing the dynamic meaning-making process for cultural colour meanings in the global market place. Grounded in the cultural approach and multimodality, and based on empirical data from Sino-Finnish business encounters, the paper presents a tripartite perspective to the provenance of cultural colour meanings: persisting colour meanings, cultural colour heritage, and transforming colour meanings.
The framework acknowledges that the cultural meanings of colour originate from shared cultural experiences related to relatively persisting colour meanings or relatively transforming colour meanings. The persisting colour meanings can be related to religious colours, ideological/value-based colours, national colours, ethnic colours, linguistic colour meanings, or colours of the environment, while the transforming colour meanings can be related to colours from “other” cultures, colours of particular eras, colours of ethnic, religious, or linguistic minorities, colour expressing wealth, gender-bound colours, or colours of younger generations. Moreover, the colours related to shared experience in reference to religious, ideological or other value-bound colours form the cultural colour heritage within a cultural context. This cultural colour heritage acts as a type of filter between the other persisting colour meanings and the transforming colour meanings in the dynamic process of negotiating and re-negotiating cultural colour meanings. The framework is visualised as a globe where various cultural experiences interact. These aspects are illustrated with rich photographic data from China and from Finland.
Finally, the paper concludes that in cross-cultural context, the communicative functions of colour are ideational, interactional, textual and expressive. Communicators use resources from their particular context-specific cultural identity and experience to assign meaning for a colour. Consequently, the communicators become expressors who, intentionally or unintentionally, express their cultural experience via colours. Therefore, in a cross-cultural context the modality of colour may gain new meanings, and “meaning-making” is not in the control of the communicator/expressor.
... In such games, depending on various conditions and the evolutionary dynamics employed, shared meaning can evolve in the sense that the actions taken by the receivers are always effective. The earliest applications that the authors are aware of involving game-theoretic evolution of the shared meaning of categories (as opposed to the game-theoretic evolution of shared meanings of individual objects) are Belpaeme's 2002 Ph.D dissertation, [5], Dowman's 2003 article [8], and the well-known 2005 article of Steels and Belpaeme [32]. These applications involve the categorization of color. ...
Linguistic meaning is a convention. This article investigates how such conventions can arise for color categories in populations of simulated "agents". The method uses concepts from evolutionary game theory: A language game where agents assign names to color patches and is played repeatedly by members of a population. The evolutionary dynamics employed make minimal assumptions about agents' perceptions and learning processes. Through various simulations it is shown that under different kinds of reasonable conditions involving outcomes of individual games, the evolutionary dynamics push populations to stationary equilibria, which can be interpreted as achieving shared population meaning systems. Optimal population agreement for meaning is characterized through a mathematical formula, and the simulations presented reveal that for a wide variety of situations, optimality is achieved.
... While both de Boer's model, and the evolutionary model of colour terms described in this thesis, aim to explain language typology using same kind of computer modelling methodology, the expression-induction model which is most relevant to the model of colour term evolution presented in this thesis is that of Belpaeme (2002). Belpaeme's model also looks at the cultural evolution of colour term systems over time 25 , although the details of the model are somewhat different from the model presented here, as are the aspects of colour term systems for which it is able to account. ...
... The model of colour term evolution presented in this thesis builds on the work of Belpaeme (2002). It uses a similar methodology, in that it is also a kind of expressioninduction model, but there are a number of key differences. ...
... Some of the findings of the psychological studies of colour perception and colour naming reviewed above were incorporated into the model, and this allowed it to account for much of the typological data concerning colour words, as will be shown below. 27 Belpaeme (2002) did suggest that the split into light and dark colours seen in languages with only two colour terms might be explainable in terms of his model, because this might be the easiest way to divide up the colour space, but, in its present form, the model would not be able to account for any other aspects of colour term typology. ...
This thesis investigates language acquisition and evolution, using the methodologies of Bayesian inference and expression-induction modelling, making specific reference to colour term typology, and syntactic acquisition. In order to test Berlin and Kay's (1969) hypothesis that the typological patterns observed in basic colour term systems are produced by a process of cultural evolution under the influence of universal aspects of human neurophysiology, an expression-induction model was created. Ten artificial people were simulated, each of which was a computational agent. These people could learn colour term denotations by generalizing from examples using Bayesian inference, and the resulting denotations had the prototype properties characteristic of basic colour terms.
... In common with the work reported here, Belpaeme (2002), Steels and Belpaeme (2005), and Belpaeme and Bleys (2005) applied the expression-induction methodology to color term evolution, although the details of each of these models were quite different from those of the model presented here. Belpaeme's (2002) simulations typically contained 10 agents, each of which was able to represent color categories using adaptive networks, a kind of neural network, allowing color categories of almost any size or shape to be represented. Color in the model was represented in terms of the CIE L*a*b* space, which was chosen because Lammens (1994) showed that his computer model of color naming worked best in that space. ...
... The most important result of Belpaeme (2002) was that, so long as communication was simulated, coherent color lexicons emerged that were shared by all the agents. The color lexicons would divide the color space into a number of color regions, each of which would be associated with a particular color word. ...
... The agents never agreed completely about the exact meaning of each word, but their languages were consistent enough for them to achieve rates of communicative success in excess of 85%. However, the color categories emerging in Belpaeme's model did not resemble the color terms of real languages, as they did not conform to the typological restrictions observed in color term systems cross-linguistically. 13 Belpaeme and Bleys (2005) reported simulations conducted using a modified version of Belpaeme's (2002) model. In the new model, each color term was represented by locating its center at a point in the color space, and using the Euclidean distance from that point as a membership function. ...
An expression-induction model was used to simulate the evolution of basic color terms to test Berlin and Kay's (1969) hypothesis that the typological patterns observed in basic color term systems are produced by a process of cultural evolution under the influence of biases resulting from the special properties of universal focal colors. Ten agents were simulated, each of which could learn color term denotations by generalizing from examples using Bayesian inference, and for which universal focal red, yellow, green, and blue were especially salient, but unevenly spaced in the perceptual color space. Conversations between these agents, in which agents would learn from one another, were simulated over several generations, and the languages emerging at the end of each simulation were investigated. The proportion of color terms of each type correlated closely with the equivalent frequencies found in the World Color Survey, and most of the emergent languages could be placed on one of the evolutionary trajectories proposed by Kay and Maffi (1999). The simulation therefore demonstrates how typological patterns can emerge as a result of learning biases acting over a period of time.
... Classifying pixels into perceptual colour categories [15] is a non-trivial task. There is a considerable body of work in both human and automated colour category naming and identification [3][16] [17]. Based on the findings in colour category theory we partitioned the HSL colour space into the colour categories of White, Black, Red, Green, Blue, Yellow, Gray, Pink, Purple, Orange and Brown. ...
In this paper we present a method for automatically evaluating the amount of colour changes images undergo when perceived
by individuals with colour deficient vision. This measure enables the classification of images based on the extent images
visually change when viewed by people with one of the three classes of dichromatic (protanopia, deuteranopia, and tritanopia)
colour vision. By measuring the extent that colour images appear perceptually different a designer, or automated layout technique,
will have an indication of whether the choice of colour usage in an image could lead to colour ambiguity or colour confusions.
