Perception-action Learning Research Articles

Emergent Intentionality in Perception-Action Subsumption Hierarchies

A cognitively-autonomous artificial agent may be defined as one able to modify both its external world-model and the framework by which it represents world, requiring two simultaneous optimization objectives. This presents deep epistemological issues centered on the question of how a framework for representation (as opposed to the entities it represents) may be objectively validated. In this summary paper, formalizing previous work in this field, it is argued that subsumptive perception-action learning uniquely has the capacity to resolve these issues by a) building the perceptual hierarchy from the bottom up so as to ground all proposed representations and b) maintaining a bijective coupling between proposed percepts and projected action possibilities to ensure empirical falsifiability of these grounded representations. In doing so, we will show that such subsumptive perception-action learners intrinsically incorporate a model for how intentionality emerges from randomized exploratory activity in the form of 'motor babbling'. Moreover, such a model of intentionality also naturally translates into a model for human-computer interfacing that makes minimal assumptions as to cognitive states.

From play to proficiency: The ontogeny of stone-tool use in coastal-foraging long-tailed macaques (Macaca fascicularis) from a comparative perception-action perspective.

Macaques crack shellfish in coastal environments with specialized stone-hammering techniques. I provide the first examination of skill development from 866 object-manipulation and 7,400 tool-use bouts, collected over 15 months, using longitudinal analyses of infants' object manipulation (N = 7) and cross-sectional comparisons of manipulative and tool-use behavior (N = 69). I adopt a Perception-action approach, examining how the emergence of actions on objects relate to the spatial-relational and percussive challenges of tool use. Infants begin manipulating single items, particularly stones, at 1-2 months. Combining objects predominates (78%) by 1.5-2.5 years, and bouts involving food and tools but with incorrect spatial relations and action sequences prevail (73%) by 2.5-3.5 years. Placing, precedes rubbing objects on surfaces. Percussion emerges last, as disorganized striking before becoming consistent and targeted. Macaques manipulate combinations of stones and oysters, before stones, anvils, and motile shellfish, but success on either food type is only observed at 2.5-3.5 years. After competence, success rates and strike accuracy improve within 3 months on oysters and 5 months on motile shellfish. Older tool users (>4.5 years) had higher success rates, strike accuracy, strike efficiency, and tool fidelity. Macaque tool-use appears facilitated by a propensity for stone manipulation, but challenged by mastering spatial relations and percussion. I relate my findings to the development of stone-tool use in capuchins and chimpanzees, stone-handling in related macaque species, object play in Old World monkeys, and percussion in children, to further understand how biological propensities, environments, and social influences contribute to perception-action learning across species. (PsycINFO Database Record

Social-motor experience and perception-action learning bring efficiency to machines.

Lake et al. proposed a way to build machines that learn as fast as people do. This can be possible only if machines follow the human processes: the perception-action loop. People perceive and act to understand new objects or to promote specific behavior to their partners. In return, the object/person provides information that induces another reaction, and so on.

Perceptuo-motor learning rate declines by half from 20s to 70/80s

This study examined perception-action learning in younger adults in their 20s compared to older adults in their 70s and 80s. The goal was to provide, for the first time, quantitative estimates of perceptuo-motor learning rates for each age group and to reveal how these learning rates change between these age groups. We used a visual coordination task in which participants are asked to learn to produce a novel-coordinated rhythmic movement. The task has been studied extensively in young adults, and the characteristics of the task are well understood. All groups showed improvement, although learning rates for those in their 70s and 80s were half the rate for those in their 20s. We consider the potential causes of these differences in learning rates by examining performance across the different coordination patterns examined as well as recent results that reveal age-related deficits in motion perception.

Learning a coordinated rhythmic movement with task-appropriate coordination feedback

A common perception-action learning task is to teach participants to produce a novel coordinated rhythmic movement, e.g. 90 degrees mean relative phase. As a general rule, people cannot produce these novel movements stably without training. This is because they are extremely poor at discriminating the perceptual information required to coordinate and control the movement, which means people require additional (augmented) feedback to learn the novel task. Extant methods (e.g. visual metronomes, Lissajous figures) work, but all involve transforming the perceptual information about the task and thus altering the perception-action task dynamic being studied. We describe and test a new method for providing online augmented coordination feedback using a neutral colour cue. This does not alter the perceptual information or the overall task dynamic, and an experiment confirms that (a) feedback is required for learning a novel coordination and (b) the new feedback method provides the necessary assistance. This task-appropriate augmented feedback therefore allows us to study the process of learning while preserving the perceptual information that constitutes a key part of the task dynamic being studied. This method is inspired by and supports a fully perception-action approach to coordinated rhythmic movement.

A linear-complexity reparameterisation strategy for the hierarchical bootstrapping of capabilities within perception–action architectures

Perception–action (PA) architectures are capable of solving a number of problems associated with artificial cognition, in particular, difficulties concerned with framing and symbol grounding. Existing PA algorithms tend to be ‘horizontal’ in the sense that learners maintain their prior percept–motor competences unchanged throughout learning. We here present a methodology for simultaneous ‘horizontal’ and ‘vertical’ perception–action learning in which there additionally exists the capability for incremental accumulation of novel percept–motor competences in a hierarchical fashion. The proposed learning mechanism commences with a set of primitive ‘innate’ capabilities and progressively modifies itself via recursive generalising of parametric spaces within the linked perceptual and motor domains so as to represent environmental affordances in maximally-compact manner. Efficient reparameterising of the percept domain is here accomplished by the exploratory elimination of dimensional redundancy and environmental context. Experimental results demonstrate that this approach exhibits an approximately linear increase in computational requirements when learning in a typical unconstrained environment, as compared with at least polynomially-increasing requirements for a classical perception–action system.

A new breed of robot

Designers of artificial cognitive systems (ACS) have, until recently, tended to adopt one of two approaches to thinking robots: classical rule-based artificial intelligence or artificial neural networks. However, a new breed of cognitive, learning robot developed through the European Union-funded project COSPAL (cognitive systems using perception-action learning) combines the best of both worlds. In this paper, the author describes the challenges faced by artificial cognitive system design and defines the progress made through the COSPAL project initiative.