BenceV / Honours-PushesAndGNNs

In this project I explore an potential approach to estimate a human’s intention in a dyadic collaborative manipulation task by learning to predict the intended future trajectory of the co-manipulated object via the latent graph representation of the system.

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Honours Project Repository

Welcome to my the main repository of my Honours Project. Here you can find almost every piece of code I have worked on throughout the entirety of the project. The code base is distributed in to 3 distinct phases:

  • Phase 1: Initial experimentations
  • Phase 2: Rigidbody experiments
  • Phase 3: Realworld Data experiments

Each of these phases can be considered a milestone in the project and are placed into their own branches inside the repository.

About

In this project I explore an potential approach to estimate a human’s intention in a dyadic collaborative manipulation task by learning to predict the intended future trajectory of the co-manipulated object via the latent graph representation of the system.