Sensorimotor Control and Learning

In the Sensorimotor Control and Learning Group we are interested in human multisensory perception and goal-oriented behaviour in virtual reality (VR) as well as the real world. When we interact with our direct surroundings, e.g. pick up a cup of coffee, we do so seemingly effortlessly. Yet, even such simple every-day tasks involve sensing the 3D position and orientation of the cup (sensory processing), making an educated guess whether it is currently full or empty to estimate its weight (cognitive processing), and having an idea of our own limb position and where it needs to go next (body perception and movement planning).

People

Teaching

Publications

Projects

The research in our group focuses on how our sensory system selects and combines relevant pieces of information for the perception of our environment, as well as for performing such goal-oriented tasks as described above. As part of this we also investigate how our senses combine to form a sense of ownership over our own body, an avatar, or a tool and a sense of agency about our actions and their effects.

To address these topics, we use straightforward perceptual tasks (psychophysics), game-like target-pointing or tracking tasks, as well as mathematical modelling approaches (e.g. Bayesian inference or Optimal Control) to approach these topics also from the theoretical side.

Below we provide an outline of some of the topics we are currently working on.

The sense of agency is a subjective feeling of having caused an event through our own action. The current understanding of when we perceive such a causal link between our action and the action outcome is still limited.

As part of the The Adaptive Mind (TAM) project we investigate the sense of agency both from a theoretical point of view, using Bayesian causal and control models and machine learning (ML) techniques, and from an empirical side, developing new methods to measure agency in motor control tasks.

Moreover, a particular focus of the TAM project is also to form an understanding of how predictive and adaptive processes are affected in psychoses, such as observed in patients with schizophrenia spectrum disorder (SSD). It is known that patients with SSD or even persons with strong schizotypal traits can have an altered perception of causality, agency, or body structure representation (Fotia, Jason, Van Dam, Ferri, Romei, 2021; Fotia, Van Dam, Sykes, Ambrosini, Costantini, Ferri, 2022).

Therefore, in cooperation with the Translational Neuroimaging Lab from Marburg we are comparing healthy control participants with schizophrenic patients. The aim of the project is to gain a better understanding of the sense of agency in general and, more particularly, how this is affected in schizophrenia.

Objectives:

  • Update the methodology for measuring agency
  • Build a Bayesian model for the sense of agency
  • Examine the potential development of the sense of agency using ML-techniques
  • Elucidate parameters that might be affected in schizophrenia

Embodiment is the process by which we relate physical entities or events to our own bodily representations. These representations can be related to our body in different ways, such as feeling that a body part or an external object is part of our own body (Ownership) or perceiving that an event in the environment was caused by our own action (Agency). A sense of Embodiment has been linked to multisensory correspondences between e.g., body-related tactile sensation and vision.

However, often paradigms rely on the participant remaining static whilst multisensory stimuli are presented in an external (non-self-induced) way. As a result, the role of motor control and the resulting body-related sensory inputs, which are key for potential areas of application (e.g., avatar immersion in VR or the use of external tools or prosthetic devices) has been largely unexplored.

The aim of this project is to fill this gap and focus on the role of action-related sensations and Agency for inducing a sense of Ownership. Particularly, the project explores the role of proprioception, action related tactile sensations, repeated exposures, and prior knowledge of body appearance. The results will be informative to help create optimal conditions for creating a sense of Embodiment for avatars in Virtual Reality, tools, and prosthetic devices.

Objectives:

  • Investigate the sense of ownership under repeated exposure to embodiment illusions
  • Investigate cue-combination during active movement for the senses of agency and ownership
  • Investigate the role of vibrotactile feedback for motor control and embodiment

When we humans interact with the world around us, we heavily rely on visual information to identify and locate objects in our direct environment and guide our actions whilst avoiding obstacles. This means that visual misperceptions are likely to influence our actions.

In this project we investigate the influence of visual patterns on our movement behaviour in the broadest sense. This includes investigating how high-contrast repetitive patterns on walls, facades, and other building constructions can influence our sense of balance on the one hand, and how illusory motion of external objects can influence target-oriented movements on the other.

Through this research we aim to both get a better understanding of our sensorimotor control system as well as provide guidelines for creating living spaces that can be navigated safely.

Objectives:

  • Compare visual information concerning position and movement of external objects and our own limbs to guide our actions.
  • Investigate balance control and gait under the influence of background motion with various visual patterns
  • Model balancing behaviour and the learning mechanisms involved using existing control models and machine learning