6 papers
PixelGen: Improving Pixel Diffusion with Perceptual Supervision
Zehong Ma, Ruihan Xu, Shiliang Zhang
Pixel diffusion generates images directly in pixel space, avoiding the VAE artifacts and representational bottlenecks of two-stage latent diffusion. Recent JiT further simplifies p…
DeCo: Frequency-Decoupled Pixel Diffusion for End-to-End Image Generation
Zehong Ma, Longhui Wei, Shuai Wang +2
Pixel diffusion aims to generate images directly in pixel space in an end-to-end fashion. This approach avoids the limitations of VAE in the two-stage latent diffusion, offering hi…
MagCache: Fast Video Generation with Magnitude-Aware Cache
Zehong Ma, Longhui Wei, Feng Wang +2
Existing acceleration techniques for video diffusion models often rely on uniform heuristics or time-embedding variants to skip timesteps and reuse cached features. These approache…
Efficient Multi-modal Long Context Learning for Training-free Adaptation
Zehong Ma, Shiliang Zhang, Longhui Wei +1
Traditional approaches to adapting multi-modal large language models (MLLMs) to new tasks have relied heavily on fine-tuning. This paper introduces Efficient Multi-Modal Long Conte…
Multi-modal Reference Learning for Fine-grained Text-to-Image Retrieval
Zehong Ma, Hao Chen, Wei Zeng +2
Fine-grained text-to-image retrieval aims to retrieve a fine-grained target image with a given text query. Existing methods typically assume that each training image is accurately…
OVMR: Open-Vocabulary Recognition with Multi-Modal References
Zehong Ma, Shiliang Zhang, Longhui Wei +1
The challenge of open-vocabulary recognition lies in the model has no clue of new categories it is applied to. Existing works have proposed different methods to embed category cues…