Invited speakers

Planning, selecting, and monitoring tool-use actions: evidence from left-hemisphere stroke

Tool use is a hallmark of human cognition, yet is the subject of comparatively little study. Research in our lab has shown that knowledge of how tool actions look and how they feel depend on distinct nodes in a distributed left hemisphere network. Lesion studies show that damage to the tool use network (and resulting limb apraxia) are associated with slow activation of semantic knowledge of actions, with specific consequences for predictive processing, selection of task-relevant actions, and using competition as a signal that monitoring is required. In addition, priming and eyetracking studies indicate that tool action representations are organized by the brain—and compete with one another-- in terms of the similarity of their sensory-motor features. Together, this work shows that the selection and monitoring of tool actions have many parallels with semantic selection and monitoring as typically studied in the context of language, and suggests a number of lines of inquiry for future work.

The development and functional contributions of objects representations in the dorsal pathway

The human visual system's capacity for object recognition and interaction hinges on two distinct cortical pathways: the ventral occipitotemporal pathway, supporting object recognition, and the dorsal occipitoparietal pathway, facilitating visually guided actions. In my talk, I will first discuss recent evidence from a series of neuropsychological, and neuroimaging studies that demonstrate the plausible role of the dorsal pathway in object perception and reveal an interplay between shape representations derived by the two pathways. Next, I will describe a series of developmental studies with typically developing children and individuals with neurodevelopmental disorders, such as amblyopia, cortical resections, and ASD that examine the development of visuomotor behaviors and the functional dissociation between perception and action. We find that object representations in the dorsal pathway are not fully matured even in school-age children, indicating a protracted developmental trajectory.

Additionally, visuomotor behaviors, which are closely linked with computations executed by the dorsal pathway, show higher susceptibility to atypical development compared to perceptual behaviors. This trend is also reflected in a reduced functional dissociation between perception and action in children with neurodevelopmental conditions. Thus, I will argue that the longer maturation period of object representations in the dorsal pathway may account for their heightened vulnerability to neurodevelopmental disorders. Collectively, these results offer a nuanced perspective on the visual system's dual pathways, underscoring the critical importance of the dorsal pathway in object perception and the complexities of its developmental dynamics.

The neural computations underlying human social interaction recognition

Humans perceive the world in rich social detail. We effortlessly recognize not only objects and people in our environment, but also social interactions between people. The ability to perceive and understand social interactions is critical for functioning in our social world. We recently identified a brain region that selectively represents others' social interactions in the posterior superior temporal sulcus (pSTS) in a manner that is distinct from other visual and social processes, like face recognition and theory of mind. However, it is unclear how social interactions are processed in the real world where they co-vary with many other sensory and social features. In the first part of my talk, I will discuss new work using naturalistic fMRI paradigms and machine learning analyses to understand how humans process social interactions in real-world settings. We find that features of a social interaction are extracted hierarchically along the STS, with strong selectivity for communicative interactions, even after controlling for the effects of other co-varying perceptual and social information. In the second part of my talk, I will discuss the computational implications of social interaction selectivity in the brain, and present a novel graph neural network model, SocialGNN, that instantiates these insights. SocialGNN reproduces human social interaction judgements in both controlled and natural videos using only visual information, without any explicit model of agents' minds or the physical world, but requires relational, graph structure and processing to do so. Together, this work suggests that social interaction recognition is a core human ability that relies on specialized, structured visual representations.

Intuitive Physics in the Brain

Meaning in Action

I will discuss visual object recognition through the lens of how "downstream" systems interact with the outputs of visual recognition processes.
Those downstream processes include conceptual interpretation, grasping and object use, navigating and orienting in an environment, physical reasoning about the world, and inferring future actions and the inner mental states of agents. I will emphasize functional neuroimaging studies in patients with acquired brain lesions to show that lesions outside the ventral visual object processing stream can cause category-specific patterns of information disruption within the ventral stream. These findings support the view that connectivity between ventral stream object recognition areas and other regions of the brain is the basis for the emergence of neural specificity for a limited number of semantic domains in the brain.

Perception from the inside out

Current models of object perception maintain that perception arises from a series of hierarchical stages proceeding from sensory input to conscious perception. That is, from the outside in. This idea runs counter to the neural architecture of perceptual systems as recurrent networks of feedforward and feedback connections. In this talk I’ll present neuropsychological and neuroimaging data in support of the idea that object perception occurs mainly from the inside out, as William James’s first proposed in 1890.

Some fMRI studies on action structure

In recent years, many neurophysiological and behavioral studies have addressed the question of how we process actions of others. A prominent role has been played by the insight that action planning, imagination, and observation activate a joint network that has come to be called the 'action observation network' or AON. Most of the studies that contributed to this discovery used minimal actions as stimulus material. However, everyday actions consist of many such minimal actions, which can be described as segments of larger action units. In my talk, I report on some fMRI studies in which we investigated the processing of this segmental structure of observed actions. The results suggest that the AON is part of a larger network for processing suprasegmental actions, and provide a link to research on episodic and script memory.

Copyright: Peter Leßmann