Zigeng Chen

SlimSAM. [NeurIPS 2024] SlimSAM: 0.1% Data Makes Segment Anything Slim

AsyncDiff. [NeurIPS 2024] AsyncDiff: Parallelizing Diffusion Models by Asynchronous Denoising

DMax. DMax: Aggressive Parallel Decoding for dLLMs

Awesome-Efficient-Segment-Anything. One summary of efficient segment anything models

CoDe. [CVPR 2025] CoDe: Collaborative Decoding Makes Visual Auto-Regressive Modeling Efficient

VeriThinker. [NeurIPS 2025] VeriThinker: Learning to Verify Makes Reasoning Model Efficient

dParallel. [ICLR 2026] dParallel: Learnable Parallel Decoding for dLLMs

NUS-EE5907-CA1-by-Chen-Zigeng. Jupyter Notebook

NUS-EE5907-CA2-by-ChenZigeng. Jupyter Notebook

chenzigeng99. academic personal homepage