Talent.com
Human Intention Understanding during Human-Human and Human-Collaborative Robot Tasks

Human Intention Understanding during Human-Human and Human-Collaborative Robot Tasks

Université de SherbrookeSherbrooke, Canada
30+ days ago
Salary
CA$120,000.00–CA$160,000.00 yearly
Job description

Topic description

The primary objective of this Ph.D. project is to investigate and interpret human intentions in the context of cooperative tasks involving both humans and robots. The project will involve conducting a series of natural human-human cooperative tasks and analogous experiments incorporating human-robot collaborations. The study aims to establish a taxonomy of cooperative tasks performed by humans and robots, identifying similarities and differences between the two types of tasks. Data from various sources, including motion capture technology and neurophysiological measurements such as EEG and EMG, will be combined, analyzed, and observed to derive insights.

The project's second contribution involves developing a model of human intention understanding. It is hypothesized that the human nervous system comprehends others' intentions through motor simulation, where observed actions are mapped to the observer's own motor representations. This hypothesis draws support from the discovery of mirror neurons in the monkey premotor cortex, which exhibit similar firing patterns when performing or observing specific actions. The study will explore the relationship between EEG oscillatory activity in the mu rhythm (8-15 Hz), mirror neuron function, and action understanding. By contrasting sensorimotor mu rhythm in scenarios where participants observe either a human or a robot performing an action, and where the intended goal of the action is simple or complex, the project aims to identify differential modulations in mu activity.

To address the challenge of processing the large volume of neurophysiological and biological data (EEG and EMG), the project will leverage machine learning techniques. Recent advancements in machine learning, including classifiers, deep learning, and generative adversarial networks (GANs), will be utilized to classify and process the collected data.

Funding category

Public / private mixed funding

Funding further details

Financement disponible