|
Joey Tianyi Zhou
As the Head of the Applied AI Systems & Agents Chapter at the Institute of Advanced Intelligence and Computing (IAIC), Agency for Science, Technology and Research (A*STAR),
I drive the development of the A*STAR Agentic AI platform. Simultaneously, as Deputy Director at the Centre for Frontier AI Research (CFAR). I am also heading AI Engineering & Platform at the National Applied AI Centre of Excellence (CoE), driving the adoption of Agentic AI across companies in Singapore.
Concurrently, I am also tenured Associate Professor of Computer Science in Singapore Management Univerisity (SMU).
I obtained my Ph.D. from Nanyang Technological University (NTU), Singapore, in 2015.
Prior to my current roles at A*STAR, I worked as a Senior Research Engineer at the Sony US Research Center in San Jose, California.
A*STAR Email / SMU Email / Github / Google Scholar / |
|
|
|
| Research |
|
Resource-efficient AI |
|
|
Beyond Modality Collapse: Representation Blending for Multimodal Dataset Distillation
|
|
|
CondenseLM: LLMs-driven Text Dataset Condensation via Reward
Matching |
|
|
Breaking Class Barriers: Efficient Dataset Distillation via Inter-Class Feature Compensator |
|
|
Diversity-Driven Synthesis: Enhancing Dataset Distillation through Directed Weight Adjustment |
|
|
Evolution-aware VAriance (EVA) Coreset Selection for Medical Image Classification |
|
|
Spanning Training Progress: Temporal Dual-Depth Scoring (TDDS) for Enhanced Dataset Pruning |
|
|
Multisize Dataset Condensation |
|
|
You Only Condense Once: Two Rules for Pruning Condensed Datasets |
|
|
Sequential Subset Matching for Dataset Distillation |
|
|
Sharpness-aware training for free |
|
|
EDCompress: Energy-Aware Model Compression for Dataflows |
|
|
Dual Adversarial Neural Transfer for Sequence Labeling |
|
|
Multi-source Meta Transfer for Low Resource Multiple Choice Question Answering |
|
|
A Deep Learning Framework for Hybrid Heterogeneous Transfer Learning |
|
|
Multi-class Heterogeneous Domain Adaptation |
|
|
Safe and Reliable AI |
|
|
TokenSwap: Backdoor Attack on the Compositional Understanding of Large Vision-Language Models |
|
|
Poisoned Distillation: Injecting Backdoors into Distilled Datasets Without Raw Data Access |
|
|
Identifying and Mitigating Social Bias Knowledge in Language Models |
|
|
Shortcuts Arising from Contrast: Towards Effective and Lightweight Clean-Label Attacks in Prompt-Based Learning |
|
|
Meta Knowledge Condensation for Federated Learning |
|
|
Trusted Multi-View Classification with Dynamic Evidential Fusion |
|
|
XAI Beyond Classification: Interpretable Neural Clustering |
|
|
PointBA: Towards Backdoor Attacks in 3D Point Cloud |
|
|
Trusted Multi-View Classification |
|
|
Trustworthy Multimodal Regression with Mixture of Normal-inverse Gamma Distributions |
|
|
Is BERT Really Robust? A Strong Baseline for Natural Language Attack on Text Classification and Entailment |
|
|
Query-efficient Meta Attack to Deep Neural Networks |
|
|
Large Foundation Model and Agentic Framework |
|
|
Evolving Quantitative Reasoning through Self-Play in Digital Twin Markets, |
|
|
Agentic Spatio-Temporal Grounding via Collaborative Reasoning, |
|
|
Rethinking Agent Design: From Top-Down Workflows to Bottom-Up Skill Evolution |
|
|
Agent Trading Arena: A Study on Numerical Understanding in LLM-Based Agents |
|
|
STADe:
Sensory Temporal Action Detection via Temporal-Spectral Representation Learning |
|
|
KPL: Training-Free Medical Knowledge Mining of Vision-Language Models |
|
|
Word2Pix: Word to Pixel Cross-Attention Transformer in Visual Grounding |
|
|
Difformer: Multi-resolutional differencing transformer with dynamic ranging for time series analysis |
|
|
Locality-Aware Crowd Counting |
|
|
Natural Language Video Localization: A Revisit in Span-based Question Answering Framework |
|
|
AnomalyNet: An Anomaly Detection Network for Video Surveillance |
|
| Competitive Funding Award |
|
|
|