Search Results

The motor theory of biological motion perception hypothesizes that motor commands (or records thereof) are used to recognize human event recognition, motor theory, biological motion perception movements when they are visually perceived. However, current theories of human action render this motor theory redundant. This chapter argues that motor commands are not responsible for the specific forms of different kinds of movements such as running or walking. Rather, passive dynamical organizations are used to generate forms of movement that are then controlled by parametrically adjusting the dynamics. However, it is the dynamically generated movement forms that can provide the information that allows biological motions to be perceived and recognized for what they are. This possibility has been systematically investigated in a number of studies inspired by an ecological approach to visual event perception. The approach hypothesizes that lawfully generated information must be available to allow perception and support recognition. Trajectory forms generated by event dynamics would provide such information. The studies have shown that trajectory forms can be used by human observers to recognize events.

This chapter begins with a discussion of the significance of events. It then focuses on four traditional object-perception issues and considers potential analogies to event perception. These are segmentation and grouping; what makes two events similar; representation; and feature binding. An event taxonomy is then presented. It is argued that events, like objects, have boundaries. These boundaries may reflect the statistical structure of events as we have experienced them, either individually or as a species.

Somewhere along the line of cognitive processes going from perception to recounting, a transition must take place by which this continuous stream of events is segmented into a sequence of discrete units. This chapter analyzes event segmentation. Darren Newtson's model of the perceptual and cognitive mechanisms of event segmentation is discussed. According to this model, observers are not confined to one invariant type of segmenting over a given course of events. Instead, the definition of event segments results from a complex interplay of both characteristics of the observed event and of personal factors, the latter ranging from stable cognitive traits to observational goals, prior knowledge, and mood states. As a consequence, variations in segmentations have been shown to have a substantial impact on higher cognitive processes such as memorization, causal attribution, decision making, or experience of time.

This chapter focuses on what children need to know about events before they learn their first verbs and, more broadly, their first relational terms. Topics covered include when infants begin to perceive, process, and represent actions and spatial relations; the cues used to segment events into meaningful units; research on infants' ability to discriminate the semantic components containment, support, and degree of fit; research on infants' discrimination of path and manner; and factors that hinder or support infants' abstraction and categorization of spatial relations.

This chapter begins with a discussion of the notion of an event. It argues that various ideas associated with the object/event distinction are not on firm ground, and considers the claim that events (as opposed to objects) are changes. It concludes that the universe does not come parceled into events, and the events individuated do not arrive with “natural” boundaries among their parts that simply await inspection, perception, recognition, and description. Thus, both events themselves and the theoretically significant concepts of “event” and “event perception” — which hope to go beyond loose, unreflective, everyday talk — also require construction.

This chapter examines how perceptual processes help segment events, begin the complex process of giving meaning to the flux of change around each of us, and ultimately give us the social graces to differentiate friend from foe. It briefly considers the scope of this problem and then addresses in detail an analysis that focuses on how object paths play a vital role in event segmentation.

This chapter examines the difficulty facing children attempting to learn novel event labels. Children must overcome what has become known as the “packaging problem”: they must figure out which event components among the many that co-occur are bundled, or “packaged”, together within the meaning of an event word. It argues that children initially use two main sources of information to help them learn event words. The first is prelinguistic universal concepts, which give them a toehold into abstracting and labeling important event features. The second is the use of perceptual similarity across same-labeled exemplars, which initially makes verb meanings quite conservative and situation-specific.

We use our knowledge of causal relationships to imagine possible events. We also use these relationships to look deep into the past and infer events that were not witnessed or to infer what can not be directly seen in the present. Knowledge of causal relationships allows us to go beyond the here and now. This chapter introduces a new theoretical framework for how this very basic concept might be mentally represented. It proposes an epistemological theory of causation — that is, a theory that specifies the nature of people's knowledge of causation, the notion of causation used in everyday language and reasoning.

Episodic memories must have beginnings and endings, but these remain unknown. This chapter presents two initial exploratory studies that attempt to identify the beginnings and endings of various recent and remote episodic memories. It also considers the nature of episodic and autobiographical memory, as well as what is known about the perception of event boundaries, which must surely play a role in the formation of memories of experienced events.

This chapter analyzes the concept of an event and of event representation as an umbrella notion. It provides an overview of different ways events have been dealt with in philosophy, linguistics, and cognitive science. This variety of positions has been construed in part as the result of different descriptive and explanatory projects. It is argued that various types of notions — common-sense, theoretically revised, scientific, and internalist psychological — be kept apart.

Event modeling systems provide a semantic interpretation of sequences of pixels that are captured by a video camera. The design of a practical system has to take into account the following three main factors: low-level preprocessing limitations, computational and storage complexity of the event model, and user interaction. The hidden Markov model (HMM) and its variants have been widely used to model both speech and video signals. Computational efficiency of the Baum-Welch and the Viterbi algorithms has been a leading reason for the popularity of the HMM. Since the objective is to detect events in video sequences that are meaningful to humans, one might want to provide space in the design loop for a user who can specify events of interest. This chapter explores this using semantic approaches that not only use features extracted from raw video streams but also incorporate metadata and ontologies of activities. It presents three approaches for applications such as event recognition: anomaly detection, temporal segmentation, and ontology evaluation. The three approaches discussed are statistical methods based on HMMs, formal grammars, and ontologies. The effectiveness of these approaches is illustrated using video sequences captured both indoors and outdoors: the indoor UCF human action dataset, the TSA airport tarmac surveillance dataset, and the bank monitoring dataset.

Event sequence knowledge is necessary for learning, planning, and performing activities of daily living. Clinical observations suggest that the prefrontal cortex (PFC) is crucial for goal-directed behavior such as carrying out plans, controlling a course of actions, or organizing everyday life routines. This chapter proposes a “representational” approach to PFC function, which assumes that the PFC (a) stores long-term memories of goal-oriented event sequence knowledge and (b) seeks to establish the format and categories according to which such information is stored. It argues that the human PFC stores a unique type of knowledge in the form of structured event complexes (SECs). SECs are representations composed of higher-order goal-oriented sequences of events that are involved in the planning and monitoring of complex behavior.

In 1973, Gunnar Johansson launched the systematic study of the perception of the human body in motion. Merging his applied interests in traffic safety and his theoretical interests in vector analysis and decomposition, Johansson produced captivating displays of people dancing and doing other ordinary activities, but with the people represented only by lights on their joints. This chapter discusses Johansson’s research and some of my own follow-up on his ideas, serving as background for ideas in this volume.

Newborn infants have a special affinity for motion. It is not surprising, therefore, that the perception of biological motion has an important role in the early development of infants. This chapter describes the development of biological motion perception across early development. The chapter provides a critical discussion about whether “limitations” in infants’ visual system may reflect attunements that adaptively orient infants toward significant others (e.g., caregivers), describes the visual behavior of infants as evidence that they are actively selecting agents as targets of their attention, and provides a review of the overlaps between adult and infant vision research.

This chapter considers the cognitive penetrability of our experiences of the durations of everyday events. It defends an account of subjective duration based in contemporary psychological and neurobiological research. It shows its philosophical adequacy by demonstrating its utility in explaining the phenomenology of duration experiences. The chapter then considers whether cognitive penetrability is a problem for these experiences. It argues that, to the contrary, the problem presupposes a relationship between perception and belief that duration perceptions and beliefs do not exhibit. Instead, the assignment of epistemic features to particular processing stages appears to answer to pragmatic needs, not psychological facts.

Slow motion images in film and video are often thought to truthfully represent events in the real world, hence they are used in action replays and courts of law as evidence. While such films preserve information about what, who, where and how often, a close look at a variety of film and video images and empirical work from the author’s laboratory demonstrate that some properties of the physical world are transformed. These transformations could be considered lies in some contexts but they can also be the basis of powerful aesthetic experiences when used artfully in film. This chapter’s three level framework of image meaning is used to organize possible psychological responses to slow motion images.

The objects of language and thought establish the granularity at which cognition operates. Granularity can vary with respect to taxonomic classification, time, and space. One might suppose that basic cognitive operations such as judgments of similarity or mental imagery would be invariant over these changes in scale, but this appears not to be the case. This chapter reviews results from three domains that show how changes in grain lead to changes in the form of cognitive operations. In judgments about object features, small and medium taxonomic scales are heavily depending on an object’s parts, but large taxonomic scales are not. In perceiving events on a small temporal scale people pay close attention to actions on individual objects, but on a larger temporal scale they pay more attention to the particular objects involved. In spatial reasoning about small objects people tend to imagine objects being moved by an external force, but when reasoning about large environments they tend to imagine themselves moving within the environment. Thus, the computational form of cognitive operations appears to depend on the taxonomic, spatial, and temporal scale of the objects of those operations.

Beginning at least with Bach (1986), semanticists have suggested that objects are formally parallel to events in the way substances are formally parallel to processes. This chapter investigates whether these parallels can be understood to reflect a shared representational format in cognition, which underlies aspects of the intuitive metaphysics of these categories. The authors of this chapter hypothesized that a way of counting (atomicity) is necessary for object and event representations, unlike for substance or process representations. Atomicity is strongly implied by plural but not mass language. The chapter investigates the language–perception interface across these domains using minimally different images and animations, designed either to encourage atomicity (‘natural’ breaks) or to discourage it (‘unnatural’ breaks). The experiments test preference for naming such stimuli with mass or count syntax. The results support Bach’s analogy in perception and highlight the formal role of atomicity in object and event representation.