Verbal and Nonverbal Processing

Top PDF Verbal and Nonverbal Processing:

Verbal and nonverbal behaviour as a basis for credibility attribution: the impact of task involvement and cognitive capacity

Verbal and nonverbal behaviour as a basis for credibility attribution: the impact of task involvement and cognitive capacity

relevant information, especially the content of a message, to develop or change an attitude. This process requires higher motivation and higher cognitive ability and capacity. Persons with lower motivation and/or lower cognitive ability/capacity use the effortless mode of processing (called peripheral route in the ELM and heuristic processing in the HSM). They use, for example, easy judgmental rules (heuristics) like “experts' statements can be trusted” or “consensus opinions are correct” to form their opinion. In general, the basic assumption of dual-process theories is that the amount of a person’s motivation and capacity causes the intensity of information processing. Overall, dual-process theories have been well tested empirically (for an overview see Bohner & Waenke, 2002; Chaiken & Trope, 1999; Eagly & Chaiken, 1993; Petty & Wegener, 1999). Outside the persuasion context, dual-process
Mehr anzeigen

40 Mehr lesen

View of Nonverbal Vocalizations as Speech: Characterizing Natural-Environment Audio from Nonverbal Individuals with Autism

View of Nonverbal Vocalizations as Speech: Characterizing Natural-Environment Audio from Nonverbal Individuals with Autism

The study of nonverbal vocalizations, such as sighs, grunts, and monosyllabic sounds, has largely revolved around the social and affective implications of these sounds within typical speech. However, for individuals who do not use any traditional speech, including those with non- or minimally verbal (nv/mv) autism, these vocalizations contain important, individual-specific affective and communicative information. This paper outlines the methodology, analysis, and technology to investigate the production, perception, and meaning of nonverbal vocalizations from nv/mv individuals in natural environments. We are developing novel signal processing and machine learning methods that will help enable augmentative communication technology, and we are producing a nonverbal vocalization dataset for public release. We hope this work will expand the scientific understanding of these exceptional individuals’ language development and the field of communication more generally.
Mehr anzeigen

4 Mehr lesen

How do implicit and explicit motives differ? The role of non-verbal versus verbal stimulus and non-declarative versus declarative response formats

How do implicit and explicit motives differ? The role of non-verbal versus verbal stimulus and non-declarative versus declarative response formats

The idea of referential processing has as yet only been tested directly by Schultheiss and Brunstein (1999, 2002) using the goal imagery task. However, although translation of verbal instructions into a nonverbal representation of the goal (Hypothesis 4a) is one possible explanation for why implicit motives predicted non-declarative and declarative measures (Hypothesis 4b) after but not without goal imagery, it is clearly not the only one. In addition to the translation of a verbal into a nonverbal representation, the imagery conditions also differed from the control conditions in that participants in the imagery conditions were led to focus extensively on the incentives and disincentives associated with the goal (e.g., they imagined beating the opponent in a speed task: “And yes - now it's you who finishes first and cries stop. You hear the other guy grunting, but you are already eager for yet another test.”). Thus, not only the form of the representation differed between imagery and control participants, but also the content of the representations. As motivationally relevant aspects of the goal were strongly emphasized and elaborated in the script, these aspects were also represented, activated, and thus accessible when working on the tasks that were used to assess motivated behavior. Maybe other aspects associated with the imagery scripts (additional emphasis given to the incentives present in the situation; instruction, motivation, and opportunity to elaborate these incentives further and to relate them to personal experience), and not the verbal versus nonverbal form of representation, was decisive for the moderating effect on implicit motives.
Mehr anzeigen

157 Mehr lesen

Global-local processing relates to spatial and verbal processing: implications for sex differences in cognition

Global-local processing relates to spatial and verbal processing: implications for sex differences in cognition

These sex differences have previously been related to men and women processing the presented stimulus materials in different ways, which has often been referred to as the use of different cognitive strategies. However, the term strategy can imply a component of awareness, i.e. participants deliberately choosing to process the stim- ulus material a certain way. As this is not necessarily the case in all of the tasks for which sex differences were described, we will use the term processing style instead. It has been demonstrated that certain processing styles are more beneficial for some task than others. Thus, sex differences in overall task performance may arise, if men and women differ in their use of said processing style (for a review see ref. 3 ). Specifically, in spatial navigation tasks, men tend to take a more allocentric perspective and use a more Euclidian approach, while women tend to take a more egocentric perspective and landmark-based approach 4 – 10 . Allocentric perspective-taking is charac-
Mehr anzeigen

9 Mehr lesen

Sex differences in verbal fluency: the role of strategies and instructions

Sex differences in verbal fluency: the role of strategies and instructions

We first performed those analyses (sections “ Neutral instructions ,” “ Effect of instruction and sex on overall verbal fluency performance (number of words produced )” and “ Effect of instruction on verbal fluency strategies ”) including both, phonemic and semantic, tasks in order to evaluate whether the type of task influenced the effects of sex or instruction on the number of words, cluster size, or number of switches. However, such a large statistical model may be underpowered with the current sample size to detect significant higher-order interactions. Furthermore, it has repeatedly been suggested that phonemic and semantic fluency tasks, while sharing some common verbal processing mechanisms, require different cognitive mech- anisms and should therefore be viewed as different cognitive tasks (e.g., Henry and Crawford 2004 ; Milner and Petrides 1984 ; Troyer et al. 1997 ). Therefore, all analyses were also performed separately for the phonemic and semantic task to address effects of sex and instruction specific to each task.
Mehr anzeigen

11 Mehr lesen

Verbal memory functioning in borderline personality disorder : neuropsychological and neuroimaging perspectives

Verbal memory functioning in borderline personality disorder : neuropsychological and neuroimaging perspectives

conditions of interest (episodic retrieval: 24-hour delayed recall of a wordlist; semantic retrieval: completing a lexical fluency task) and a low level baseline (listening to MRI noise) in 18 female right-handed BPD patients and 18 non-psychiatric control subjects matched with respect to sex, age, and education. It was hypothesized that BPD patients would show increased regional BOLD responses in prefrontal and limbic brain areas during both memory retrieval conditions. Although BPD patients and control subjects showed comparable performances in verbal episodic and semantic retrieval, important group differences in regional brain activation became evident. During the retrieval of episodic information, BPD patients showed patterns of increased task-specific regional BOLD responses as compared to controls in the posterior cingulate cortex (BA 23, 31) bilaterally, in the left middle (BA 21) and superior temporal (BA 22) gyri, in the right inferior frontal gyrus (BA 45) and in the right angular gyrus (BA 39). Further, control subjects compared with BPD patients did not show areas with increased BOLD responses. During the retrieval of semantic information, BPD patients as compared with control subjects showed areas of task-specific BOLD responses with respect to the right posterior cingulate cortex (BA 31), right fusiform gyrus (BA 37), left postcentral gyrus (BA 1,2,3) and the left anterior cingulate cortex (BA 24). Again, no areas of increased task-specific BOLD responses of control subjects compared with BPD patients could be found. Despite similar neuropsychological performances of BPD patients and control subjects in episodic and semantic memory tasks the day prior to scanning, the BPD patients showed, as hypothesized, patterns of increased brain activation. However, against the hypotheses, increased regional brain activation was not only evident in prefrontal and limbic brain areas but included further parietal areas. The increased regional brain suggests that BPD patients need to recruit additional cortical resources in order to successfully retrieve information. Thus, increased activation of BPD patients during retrieval might serve as compensation (“cognitive reserve capacity”) to perform on a high level comparable to controls. Therefore, increased activation might indicate additional networks for adequate retrieval needed by BPD patients, i.e. increased effort, attention, working memory, or emotional control. However, it has to be noted that the results of the brain imaging study are limited to female patients with BPD since no male were included.
Mehr anzeigen

103 Mehr lesen

Automatic estimation of users' verbal intelligence

Automatic estimation of users' verbal intelligence

One of the main properties of a reliable dialogue system is the high quality of speech recognition. If a system is not able to completely understand certain phrases or sentences, such a communication will be frustrating and annoying for individuals. Moreover, each module of an SDS should perform its tasks at high speed in order to give the user an appropriate answer or to construct informative utterances. Ideally, users should have the feeling that they are communicating with a real dialogue partner and may express their needs and preferences using any words and expressions. Adaptive dialogue systems should constantly analyse the verbal and communication behaviour of individuals and be ready to change the dialogue strategy at any moment. They may determine the age and gender of speakers, estimate whether users have enough knowledge about the topic of interactions and avoid explaining unnecessary details. An adaptive system may allow a user to take the initiative in their conversations or to actively ask questions in order to keep the dialogue going. When recognizing the negative emotions of a user, an SDS may understand that a user is not happy with its functioning and switch to another dialogue strategy at the right time.
Mehr anzeigen

162 Mehr lesen

Agent, causer and instrument PPs in Greek : implications for verbal structure

Agent, causer and instrument PPs in Greek : implications for verbal structure

We argued that all of these roots combine with vCAUS. Internal and external causation features of the root influence the combinations of roots with particular types of Voice heads. Specifically, cause unspecified verbs alternate in all languages. Agentive and internally caused verbs don't alternate. The former only form transitive-passive constructions, the latter are generally 3 intransitive. But languages differ as to the behavior of externally caused verbs. In Greek they form anticausatives, in English only passives. Here we illustrate this difference with ‘kill’ (similarly for hit and destroy). Intransitive kill in Greek forms the anticausative (compatible with cause-PP) but not the passive (incompatible with agent-PP), see (12b), unlike English (12d). Section 2 below further discusses anticausativization of externally caused verbs in Greek. (12) a. O Janis/ o sismos/i vovma skotose
Mehr anzeigen

16 Mehr lesen

Acquiring qualitative and quantitative knowledge from verbal statements and dialogues in probabilistic domains

Acquiring qualitative and quantitative knowledge from verbal statements and dialogues in probabilistic domains

This work is part of the developnent of an intelligent envimnment, MEDICUS (Modelling, gxplanarion, and diagnostic support for somplex. qncertain lubject matteft), that [r]

10 Mehr lesen

Modellierung verbal repräsentierter Geoinformation im Anwendungsbeispiel des Katastrophenmanagements

Modellierung verbal repräsentierter Geoinformation im Anwendungsbeispiel des Katastrophenmanagements

Methodik zur Überführung verbal repräsentierter räumlicher Zusammenhänge in eine formale Darstellung vor- gestellt. Diese umfasst neben einer funktionalen, ebenfalls eine semantische Modellierungsebene. Begründet wird dies durch die differenzierten Betrachtungen, welche für eine zuverlässige und robuste Verarbeitung un- scharfer Geoinformation notwendig sind. So erfordern die verschiedenen Abstraktionsstufen der verbalen Re- präsentation semantische Modellierungsstrukturen. Diese erlauben es ebenfalls, fehlende Information durch einschränkende Bedingungen sowie heuristische Annahmen zu ergänzen. Ferner wird die der Information zu- grunde liegende Raumvorstellung, sowie die Semantik der zur Beschreibung genutzten linguistischen Terme einbezogen. Die Berücksichtigung der semantischen und topologischen Objektrelationen bei der Modellbildung ermöglicht es indes, die sprachlich bedingten semantischen Mehrdeutigkeiten aufzulösen. Formalisiert werden diese Aspekte in einer ontologischen Wissensbasis, unter Nutzung der Web-Ontologie Sprache, welche es in Form einer Beschreibungslogik erlaubt, eine konkrete Domäne semantisch korrekt abzubilden. Die Nutzung einer Inferenzmaschine gestattet es, ausgehend von dem modellierten Wissen, erforderliche Zusammenhänge abzuleiten.
Mehr anzeigen

98 Mehr lesen

View of Posed and spontaneous nonverbal vocalizations of positive emotions: Acoustic analysis and perceptual judgments

View of Posed and spontaneous nonverbal vocalizations of positive emotions: Acoustic analysis and perceptual judgments

Recognition levels per positive emotion were different for posed and spontaneous vocal expressions. There is unlikely to be single explanation for these differences, but differences in recording quality, emotion intensity, or cultural effects may have contributed. Moreover, the spontaneous vocalizations were selected by two judges in our study and it cannot be comprehensively ruled out that their selection could have been influenced by some degree of implicit bias. For instance, it is possible that the judges selected prototypical vocalizations of emotions, even though they were specifically instructed to choose vocalizations based only on the relevant context and definition. Furthermore, the intense expressions in the present study may differ from many emotional expressions that occur in daily life. Nevertheless, these results provide evidence that naïve listeners can recognize 13 different positive emotions from posed and spontaneous vocalizations.
Mehr anzeigen

6 Mehr lesen

Verbal communication at a stand-alone computer

Verbal communication at a stand-alone computer

The second focus area identifi ed by Herheim (2010) that is of relevance for this article concerns communication characteristics. Typical communicative as- pects that are highlighted as being important for pupils’ learning are thinking aloud (Kieran, 2001; Monaghan, 2005) and shared decision making (Healy, Pozzi, & Hoyles, 1995). Many researchers, e.g. Alrø and Skovsmose (2002), emphasize verbalization: to share information, to pay attention to each other’s perspectives, and justify, challenge and evaluate these perspectives. Researchers like Teasley and Roschelle (1993) and Healy et al. (1995) were among the fi rst to accentuate the im- portance of verbalization and building and maintaining channels of communication in computer settings. Wegerif (1996b) investigates how keywords can be indicators of exploratory talk. This article discusses the importance of the word wait. Wegerif (e.g. 1996a) integrates the aspect of discussion in IRF communication structures. His research context is pupil dyads or triads and their communication at comput- ers, and the D in his IDRF-structure is the part where pupils discuss a matter rath- er than giving an immediate response to the computer’s initiative. The six commu- nication patterns identifi ed and discussed in this article are all situated within this discussion aspect introduced by Wegerif.
Mehr anzeigen

28 Mehr lesen

Multimodal joke presentation for social robots based on natural-language generation and nonverbal behaviors

Multimodal joke presentation for social robots based on natural-language generation and nonverbal behaviors

Natural Language Generation (NLG) is an essential tool for provid- ing a natural human-robot interaction experience [7]. It opens up the ability to react and generate content dynamically during con- versations. In addition to the flexbility with respect to the content itself, a generation approach allows to control the robot’s linguistic style, its use of figures of speech, stylistic instruments and other language tools – characteristics, which contribute to the perceived personality. One important aspect closely linked to language and personality is humor. It regulates conversations, increases inter- personal attraction and trust in human-human interactions. In the context of conversational agents, it makes interactions more natural, enjoyable and increases credibility and acceptance [15]. For embod- ied agents, such as social robots, humorous contents are typically scripted, such as in [11–13, 27]. It is desirable to use a generative approach for generating humorous contents during interaction and to address the variety of humor. Many experiments for text-based generation have already been made, including the STANDUP [14] generator for punning riddles. However, text alone is not sufficient for a robot’s convincing humor presentation. Appropriate paralin- guistic and nonverbal behavior must be added to the text, taking facial expression, gaze, gestures and laughter into account. In this manner, multimodal generation and expression of irony based on NLG and tailored behavior has recently been explored for a Reeti robot [22], where linguistic, prosodic and nonverbal markers help the human to identify the robot’s use of irony.
Mehr anzeigen

3 Mehr lesen

Modellierung verbal repräsentierter Geoinformation im Anwendungsbeispiel des Katastrophenmanagements

Modellierung verbal repräsentierter Geoinformation im Anwendungsbeispiel des Katastrophenmanagements

Methodik zur Überführung verbal repräsentierter räumlicher Zusammenhänge in eine formale Darstellung vor- gestellt. Diese umfasst neben einer funktionalen, ebenfalls eine semantische Modellierungsebene. Begründet wird dies durch die differenzierten Betrachtungen, welche für eine zuverlässige und robuste Verarbeitung un- scharfer Geoinformation notwendig sind. So erfordern die verschiedenen Abstraktionsstufen der verbalen Re- präsentation semantische Modellierungsstrukturen. Diese erlauben es ebenfalls, fehlende Information durch einschränkende Bedingungen sowie heuristische Annahmen zu ergänzen. Ferner wird die der Information zu- grunde liegende Raumvorstellung, sowie die Semantik der zur Beschreibung genutzten linguistischen Terme einbezogen. Die Berücksichtigung der semantischen und topologischen Objektrelationen bei der Modellbildung ermöglicht es indes, die sprachlich bedingten semantischen Mehrdeutigkeiten aufzulösen. Formalisiert werden diese Aspekte in einer ontologischen Wissensbasis, unter Nutzung der Web-Ontologie Sprache, welche es in Form einer Beschreibungslogik erlaubt, eine konkrete Domäne semantisch korrekt abzubilden. Die Nutzung einer Inferenzmaschine gestattet es, ausgehend von dem modellierten Wissen, erforderliche Zusammenhänge abzuleiten.
Mehr anzeigen

100 Mehr lesen

Ricote, Sancho y los peregrinos (Don Quijote II, 54): comunicación verbal y no verbal en los encuentros interculturales

Ricote, Sancho y los peregrinos (Don Quijote II, 54): comunicación verbal y no verbal en los encuentros interculturales

En la España de la época de Cervantes los encuentros, voluntarios e involuntarios, con gente de otras culturas y de otras lenguas están al orden del día: el encuentro cotidiano entre los habitantes de España de origen cristiano, judío o mahometano, el encuentro con peregrinos extranjeros, y muchos más. Fuera de España, se producen también estos encuentros como consecuencia de los descubrimientos y conquistas, en la evangelización de pueblos no cristianos, en viajes comerciales, cruzadas o durante el cautiverio. Por ejemplo, el jesuita español José de Acosta, al describir las numerosas dificultades que surgieron durante la evangelización, menciona la gran diversidad de lenguas. Observa que tampoco una lingua franca indígena, como la que existía en los antiguos reinos incas, era suficiente para hacer comprensibles los m ysteña Fidei (Acosta 1954: 398-399, 414-415, 518-520). Pero incluso en la comprensión más rudimentaria surgían grandes problemas que fueron descritos por descubridores como Vasco da Gama, Colón o A lvar Núñez Cabeza de Vaca. Casi siempre se recurría al lenguaje corporal por señas, una forma de comunicación que, según las crónicas, funcionaba bastante bien, en cualquier caso, sin embargo, en los primeros contactos resultaba más eficaz que la comprensión verbal.1
Mehr anzeigen

19 Mehr lesen

Exascale Pre and Post Processing

Exascale Pre and Post Processing

• Objectives: development of pre and post processing as well as remote hybrid rendering tools to support exascale applications and users including:.. • Mesh manipulation and partition[r]

25 Mehr lesen

Modelling and Processing Wordnets in OWL

Modelling and Processing Wordnets in OWL

The goal of this paper is to discuss and compare different options to represent the basic entities and relations of wordnets using OWL. For this purpose, we developed alternative OWL models on the same wordnet-style resources: (1) a subset of the German wordnet GermaNet 2 , (2) the wordnet-style thesaurus TermNet 3 represent- ing technical terms from the domains of text-technology and hypertext research, and (3) the GermaTermNet representing relations between TermNet technical terms and GermaNet synsets. In Section 2, we briefly describe these resources; in Section 3 we present and compare the alternative OWL models that we created for these re- sources. The main difference between these models lies in the ontological status of the two main entity types of wordnets: the lexical units (LU) (words) and the synsets (collections of synonymous or near-synonymous lexical units). While the instance models of our resources represent LUs and synsets as individuals, the class mod- els represent these entities as classes. Class and instance models are in the scope of OWL DL and may thus be processed by description logic-based reasoners. As a third alternative, we developed metaclass models which define the two entity types LexicalUnit and Synset as metaclasses. The individuals of these metaclasses – the particular lexical units and synsets – are classes themselves, i.e. these models com- bine both the instance and the class perspective. However, the metaclass models are outside the scope of OWL DL and can thus not be processed by DL-based reasoners. Clearly, the comparison and evaluation of the models has to refer to their de- ployment in concrete applications. In Section 4, we report the results of some ex- periments in the application context of text-technological information processing: in 4.1 we compare the performance of the three OWL models in the context of au- tomatic hyperlinking using the DL reasoner RACER Pro and the Thea OWL library for SWI Prolog. In 4.2 we used the latter to calculate semantic relatedness measures on the GermaTermNet metaclass model. In the same section, we test the perfor- mance and feasability of our plug-in-approach that connects the general language resource GermaNet with the domain-specific TermNet. On the basis of the results reported in Section 4 we discuss the advantages and drawbacks of the three model types in the application contexts 4 .
Mehr anzeigen

31 Mehr lesen

Visual versus verbal elements in service advertising: the effect of endorser's facial expression and verbal content's tangibility on consumer responses / submitted by Thomas Christian Innerhofer

Visual versus verbal elements in service advertising: the effect of endorser's facial expression and verbal content's tangibility on consumer responses / submitted by Thomas Christian Innerhofer

consumers can objectively evaluate the information and better judge the quality and value of the service before they have even used it (Stafford & Day, 1995; Morgan & Reichert, 1999). Especially in restaurants or services that are evaluated by experience-based characteristics, detailed information about what the service provider offers leads to a more positive consumer attitude toward the ad than if the verbal content of an advertisement contains intangible cues (Stafford, 1996; Laskey, Seaton & Nicholls, 1994). The results of this study show that the participants of this study developed an equally positive attitude toward the ad through the verbal content with tangible and intangible cues. The non-significant main effect of the influence of the verbal content on the attitude toward the ad can be explained by the additional finding of the superiority of the visual advertising element. The additional finding shows that there is no difference in the participants' attitude toward the ad with regard to the main effect of the smiling endorser and the combination of smiling endorsers/tangible verbal cues and smiling endorsers/ intangible verbal cues. Looking at the scenarios with the non-smiling endorser, however, the verbal content with tangible cues leads to a significantly more positive attitude of the participants toward the ad than the verbal content with intangible cues. This is an indication that the visual advertising element is perceived by the participants first, and the evaluation of this element is decisive for whether the verbal advertising element is considered at all for the development of the attitude toward the ad. This can be explained by the findings of Houston, Childers and Heckler (1987), who found that the visual element of an advertisement is the first to be perceived by consumers and that the perception of this element determines how consumers process further information in the advertisement. Likewise, Bargh (2002) and Kulczynski, Ilicic and Baxter (2016) argue that when consumers see an endorser with an emotional facial expression in the advertisement, the further perception of the advertisement can be influenced to some extent. Based on this finding, the author of this thesis suggests that the assumptions of Stafford (1996) and Laskey, Seaton and Nicholls (1994) that verbal content with tangibles cues leads to a more positive attitude toward the ad are true to a certain extent, but this effect depends on how consumers perceive the visual advertising element.
Mehr anzeigen

101 Mehr lesen

Distributed and sparse signal processing

Distributed and sparse signal processing

becomes too large and the performance of the centroid estimate over the mismatched domain P t is inferior. We present the average compu- tation times of the proposed method and compared it to the (highly optimized) domain decomposition using Qhull and direct Monte-Carlo integration [LZZ07] in Tab. 5.1. As can be seen from the results, the proposed approach quickly becomes unfavorable when the number of measurements exceeds M = 5 due to the undesired scaling of com- plexity. On the positive side, the scaling does not depend on the di- mension D = N − M of the polytope so that the centroid can still be computed, e.g., in dimensions (M, N ) = (1, 100) where D = 99 and there is no hope for solutions using traditional domain decomposition or Monte Carlo integration methods. We note that we exclude  1 - and  2 -minimization for the timing analysis as they are based on a differ- ent problem formulation and do not provide estimates of volume and moments.
Mehr anzeigen

146 Mehr lesen

The attentive robot companion: learning spatial information from observation and verbal interaction

The attentive robot companion: learning spatial information from observation and verbal interaction

3.3.2. Registering a Scene Model with the Current View In order to utilize previously gathered information on a mobile robot, the system must be able to integrate multiple scene models from different view- points. In the following I will present a method to reliably fuse multiple scene models from different locations into a new model representing the current view. The method only transfers previously generated background models to the new scene, because non-static parts can be calculated if the background is known. Since the method is designed for a mobile robot, I assume that a position estimate of the camera in global coordinates is available, which drastically reduces the search space for the registration process. The Iterative Closest Point (ICP) method is used in combination with the localization information for registration of the corresponding point clouds . One of the reasons for registration is to compensate inaccuracies in the localization, which also argues for matching only the static parts to the currently visible scene. Because of the movable or even dynamic nature of the remaining parts it is likely that their location has changed since the last observation. This would make the correct registration of point clouds very hard, because registration algorithms usually try to find the best matching of the complete scene. When trying to register two point clouds that origi- nate from two very different scene configurations, the probability for making mistakes in the matching is very high. It is much saver to try to match a reduced representation that only contains the static parts of a scene to a full scene with additional movable objects than trying to match two very distinct scenes (see Figure 3.5). Of cause, the higher the confidence that the static labeled structures are in fact static, the better are the chances to correctly register scenes. Newly generated scene models with only few ob- servations involve the risk of containing actually movable structures in the static layer and therefore are endangered to suffer from major scene changes while registering scenes.
Mehr anzeigen

246 Mehr lesen

Show all 10000 documents...