Institute for Infocomm Research, A*STAR
Author pages are created from data sourced from our academic publisher partnerships and public sources.
Share This Author
A Survey of Embodied AI: From Simulators to Research Tasks
- Jiafei Duan, Samson Yu, Tangyao Li, Huaiyu Zhu, Cheston Tan
- Computer ScienceIEEE Transactions on Emerging Topics in…
- 8 March 2021
An encyclopedic survey of the three main research tasks in embodied AI – visual exploration, visual navigation and embodied question answering – covering the state-of-the-art approaches, evaluation metrics and datasets is surveyed.
ActioNet: An Interactive End-To-End Platform For Task-Based Data Collection And Augmentation In 3D Environment
- Jiafei Duan, Samson Yu, Hui Li Tan, Cheston Tan
- Computer ScienceIEEE International Conference on Image Processing…
This paper presents ActioNet, an interactive end-to-end platform for data collection and augmentation of task-based dataset in 3D environment, and the accompanying dataset is the largest task- based dataset of such comprehensive nature.
PIP: Physical Interaction Prediction via Mental Imagery with Span Selection
This work proposes a novel PIP scheme: Physical Interaction Prediction via Mental Imagery with Span Selection, which outperforms baselines and human performance in physical interaction prediction for both seen and unseen objects.
SPACE: A Simulator for Physical Interactions and Causal Learning in 3D Environments
- Jiafei Duan, Samson Yu, Cheston Tan
- Computer ScienceIEEE/CVF International Conference on Computer…
- 13 August 2021
The SPACE simulator allows the SPACE dataset, a synthetic video dataset in a 3D environment, to be generated and evaluated with a state-of-the-art physics-based deep model and shows that the SPACE datasets improves the learning of intuitive physics with an approach inspired by curriculum learning.
A Benchmark for Modeling Violation-of-Expectation in Physical Reasoning Across Event Categories
This work established a benchmark to study physical reasoning by curating a novel large-scale synthetic 3D VoE dataset armed with ground-truth heuristic labels of causally relevant features and rules and proposed the Object File Physical Reasoning Network (OFPR-Net) which exploits the dataset’s novel heuristics to outperform the authors' baseline and ablation models.
A Survey on Machine Learning Approaches for Modelling Intuitive Physics
- Jiafei Duan, Arijit Dasgupta, Jason Fischer, Cheston Tan
- Education, Computer ScienceArXiv
- 14 February 2022
A comprehensive survey of recent advances and techniques in intuitive physics-inspired deep learning approaches for physical reasoning is presented and the challenges of the current and future research directions are highlighted.
AVoE: A Synthetic 3D Dataset on Understanding Violation of Expectation for Artificial Cognition
This work proposes AVoE: a synthetic 3D VoE-based dataset that presents stimuli from multiple novel sub-categories for five event categories of physical reasoning, armed with ground-truth labels of abstract features and rules augmented to vision data, paving the way for high-level symbolic predictions in physical reasoning tasks.
PIP: Physical Interaction Prediction via Mental Simulation with Span Selection
The experiments show that PIP outperforms human, baseline, and related intuitive physics models that utilize mental simulation, and PIP’s span selection module effectively identifies the frames indicating key physical interactions among objects, allowing for added interpretability.