no code implementations • SIGDIAL (ACL) 2020 • Maryam Zare, Ali Ayub, Aishan Liu, Sweekar Sudhakara, Alan Wagner, Rebecca Passonneau
During policy training, we control for the simulated dialogue partner’s level of informativeness in responding to questions.
no code implementations • 6 Mar 2024 • Ali Ayub, Chrystopher Nehaniv, Kerstin Dautenhahn
Our results demonstrate the effectiveness of our architecture to allow a physical robot to continually adapt to the changes in the environment from limited data provided by the users (experimenters), and use the learned knowledge to perform object fetching tasks.
1 code implementation • 31 Jul 2023 • Ali Ayub, Alan R. Wagner
For most real-world applications, robots need to adapt and learn continually with limited data in their environments.
1 code implementation • 5 Jul 2023 • Christopher McClurg, Ali Ayub, Harsh Tyagi, Sarah M. Rajtmajer, Alan R. Wagner
We term this model Few-shot Incremental Active class SeleCtiOn (FIASco).
1 code implementation • 30 Jun 2023 • Ali Ayub, Jainish Mehta, Zachary De Francesco, Patrick Holthaus, Kerstin Dautenhahn, Chrystopher L. Nehaniv
Continual learning (CL) has emerged as an important avenue of research in recent years, at the intersection of Machine Learning (ML) and Human-Robot Interaction (HRI), to allow robots to continually learn in their environments over long-term interactions with humans.
no code implementations • 30 Jun 2023 • Ali Ayub, Chrystopher L. Nehaniv, Kerstin Dautenhahn
In this paper, we present a cognitive architecture for a household assistive robot that can learn personalized breakfast options from its users and then use the learned knowledge to set up a table for breakfast.
1 code implementation • 22 May 2023 • Ali Ayub, Zachary De Francesco, Patrick Holthaus, Chrystopher L. Nehaniv, Kerstin Dautenhahn
Our results suggest that participants' perceptions of trust, competence, and usability of a continual learning robot significantly decrease over multiple sessions if the robot forgets previously learned objects.
no code implementations • 9 Oct 2022 • Ali Ayub, Carter Fendley
The results show that our approach not only produces state-of-the-art results on the dataset but also allows a real robot to continually learn unseen objects in a real environment with limited labeling supervision provided by its user.
no code implementations • 19 Jul 2022 • Ali Ayub, Chrystopher L. Nehaniv, Kerstin Dautenhahn
The robot can also use the learned knowledge to correctly predict missing items over multiple weeks and it is robust against sensory and perceptual errors.
2 code implementations • 23 Mar 2021 • Ali Ayub, Alan R. Wagner
To fill this gap, we present a new dataset termed F-SIOL-310 (Few-Shot Incremental Object Learning) which is specifically captured for testing few-shot incremental object learning capability for robotic vision.
no code implementations • 13 Mar 2021 • Ali Ayub, Alan R. Wagner
Children learn continually by asking questions about the concepts they are most curious about.
no code implementations • 26 Jan 2021 • Ali Ayub, Alan R. Wagner
For many real-world robotics applications, robots need to continually adapt and learn new concepts.
1 code implementation • ICLR 2021 • Ali Ayub, Alan R. Wagner
The two main impediments to continual learning are catastrophic forgetting and memory limitations on the storage of data.
no code implementations • 22 Aug 2020 • Ali Ayub, Alan R. Wagner
The paper utilizes a recent state-of-the-art approach for incremental learning and adapts it for online learning of scenes (contexts).
no code implementations • 15 Jul 2020 • Ali Ayub, Alan R. Wagner
For many applications, robots will need to be incrementally trained to recognize the specific objects needed for an application.
no code implementations • ICML Workshop LifelongML 2020 • Ali Ayub, Alan R. Wagner
The two main challenges faced by continual learning approaches are catastrophic forgetting and memory limitations on the storage of data.
1 code implementation • 27 Feb 2020 • Ali Ayub, Alan Wagner
To solve this problem, we propose a novel approach inspired by the concept learning model of the hippocampus and the neocortex that represents each image class as centroids and does not suffer from catastrophic forgetting.
no code implementations • 3 Jan 2020 • Ali Ayub, Alan R. Wagner
The paper demonstrates a method for teaching a robot the win conditions of the game Connect Four and its variants using a single demonstration and a few trial examples with a question and answer session led by the robot.
1 code implementation • BMVC 2020 • Ali Ayub, Alan R. Wagner
Inspection of the centroids generated by our approach on RGB-D datasets leads us to propose a method for merging conceptually similar categories, resulting in improved accuracy for all approaches.