This is your work, valued

ZehongMa

Advanced
@Zehong-Ma

PH.D. student at Peking University.

zehong-ma.github.io

MagCache. The official code for NeurIPS 2025 "MagCache: Fast Video Generation with Magnitude-Aware Cache"

275

ComfyUI-MagCache. The official code that integrates MagCache (Fast Video Generation with Magnitude-Aware Cache) with ComfyUI.

274

PixelGen. Official repository for “PixelGen: Improving Pixel Diffusion with Perceptual Loss”

270

DeCo. [CVPR2026 Highlight] Official repository for “DeCo: Frequency-Decoupled Pixel Diffusion for End-to-End Image Generation”

229

OVMR. OVMR: Open-Vocabulary Recognition with Multi-Modal References (CVPR24)

36

EMLoC. Official repository for Efficient Multi-modal Long Context Learning for Training-free Adaptation (ICML 2025)

19

nwpu_19_advanced_vision. a easy project and a try for nwpu 19 vision

7

ChID-CL. Chinese idiom prediction based on contrastive learning and pretrained RoBERTa

2

SwinBERT-Extension. Python

2

OmniGen2. OmniGen2: Exploration to Advanced Multimodal Generation.

1

nwpu_20_basketball_robot. vision code of basketball robot for 2020 China Robotics Competition

1

TVFR-for-text-based-person-search. TensorFlow implementation of Text-guided Visual Feature Refinement for Text-Based Person Search accepted by ICMR 2021.

1