Dataset generation of multi-material deformable objects | |
Current state-of-the-art pipelines for Mesh Reconstruction from Point clouds for deformable objects rely on synthetically generated datasets of objects that are deformed using warping fields. While such augmentations are helpful to assess mesh reconstruction performance of methods, it is not possible to learn more about the objects beyond their abstract mesh structure. For humans, it is trivial to conclude various properties of objects based on how they behave under different forces acting on such objects; e.g., applying local force on an object. A pipeline that generates object deformations based on material properties would implicitly encode such object properties in the resulting time-series mesh deformations. Therefore, allowing for a first-of-its-kind dataset to train machine learning methods that can generalize to gain the same insight as humans about object properties. | |
Towards AI Safety: Adversarial Attack & Defense on Neural Controllers | |
The project is collaborating between SRI and RSL/CRL lab and aims to investigate the weakness of the neural controller based on the state-of-the-art [3] attacking method. | |
Learning Diverse Adversaries to Black-box Learning-based Controller for Quadruped Robots | |
The project aims to leverage the latest unsupervised skill discovery techniques to validate the state-of-the-art black-box learning-based controllers in diverse ways. | |
Towards interpretable learning pipeline: A visual-assisted workflow for locomotion learning | |
Current reinforcement learning (RL)-based locomotion controllers have shown promising performance. However we are still not clear about what is learned during the training process. In this project, we investigate the proper metrics and visualisation techniques to interactively steer the locomotion learning tasks. |
Powered by SiROP - the academic career network