Songhua Liu

EditMGT: Unleashing Potentials of Masked Generative Transformers in Image Editing

Flash-DD: An Ultra Parameter-Efficient Approach to Dataset Distillation

Paying More Attention to Images in Large-scale Dataset Compression

Understanding Dataset Distillation via Spectral Filtering

CoDA: From Text-to-Image Diffusion Models to Training-Free Dataset Distillation

SkipVAR: Accelerating Visual Autoregressive Modeling via Adaptive Frequency-Aware Skipping

Revisit Model Adaptation from Parameters to Features

Privacy-Aware Hybrid Image Synthesis with Local-Cloud Collaboration

SCoT: Self-Correction at Test-time for Image Generation

Sketch First, Scale Fast: On Efficient Inference-Time Scaling for Visual Generation

Ultra-Resolution Adaptation with Ease

Image Editing As Programs with Diffusion Models

CLEAR: Conv-Like Linearization Revs Pre-Trained Diffusion Transformers Up

LinFusion: 1 GPU, 1 Minute, 16K Image

Control and Realism: Best of Both Worlds in Layout-to-Image without Training

Heavy Labels Out! Dataset Distillation with Label Space Lightening

Video-Infinity: Distributed Long Video Generation

Injecting Vision Language into Autoregressive Image Generation