Scholar
Renjie Pi
Google Scholar ID: XUq0HwcAAAAJ
HKUST
Citations & Impact
Citations
1,813
H-index
23
i10-index
29
Publications
20
Co-authors
0
Publications
Resume (English only)
Academic Achievements
- PerceptionGPT: Effectively Fusing Visual Perception into LLM, CVPR 2024
- DetGPT: Detect What You Need Via Reasoning, EMNLP 2023
- Speciality vs generality: An empirical study on catastrophic forgetting in fine-tuning foundation models, Preprint
- Effective Bilevel Optimization via Minimax Reformulation, Preprint
- DynaFed: Tackling Client Data Heterogeneity with Global Dynamics, CVPR 2023
- Robust Federated Learning against both Data Heterogeneity and Poisoning Attack via Aggregation Optimization, Preprint
- A Holistic View of Noise Transition Matrix in Deep Learning and Beyond, ICLR 2023, Spotlight
- Self-Guided Noise-Free Data Generation for Efficient Zero-Shot Learning, ICLR 2023, Spotlight
- Measure the Predictive Heterogeneity, ICLR 2023
- Probabilistic Bilevel Coreset Selection, ICML 2022
- Model Agnostic Sample Reweighting for Out-of-Distribution Learning, ICML 2022
- G-DetKD: Towards General Distillation Framework for Object Detectors via Contrastive and Semantic-guided Feature Imitation
Research Experience
- Currently pursuing a PhD at HKUST, focusing on Data-centric AI, Multi-modal Large Language Models (MLLM), and optimization algorithms.
Background
- Currently a 3rd year PhD student in the CSE department of Hong Kong University of Science and Technology (HKUST), co-supervised by Professor Tong Zhang and Professor Xiaofang Zhou. Before that, an undergraduate student at the University of Hong Kong (HKU) majoring in Computer Engineering. Research interests mainly lie in Data-centric AI, Multi-modal Large Language Model (MLLM), and optimization algorithms.
Miscellany
- In leisure time, enjoys playing basketball, jumping rope, and jogging. Also a music lover.
Co-authors