Zhe Lin

Lavida-O: Elastic Large Masked Diffusion Models for Unified Multimodal Understanding and Generation

EditVerse: Unifying Image and Video Editing and Generation with In-Context Learning

Rethinking Global Text Conditioning in Diffusion Transformers

Frame Guidance: Training-Free Guidance for Frame-Level Control in Video Diffusion Models

Image Tokenizer Needs Post-Training

OmniClear: Soft Effects Removal from Images within a Versatile Model

Multimodal Generative Composition Recommendation

OmniVCus: Feedforward Subject-driven Video Customization with Multimodal Control Conditions

PixPerfect: Seamless Latent Diffusion Local Editing with Discriminative Pixel-Space Refinement

LaViDa: A Large Diffusion Model for Vision-Language Understanding

ImageFolder: Autoregressive Image Generation with Folded Tokens

Refine-by-Align: Reference-Guided Artifacts Refinement through Semantic Alignment

LayerFusion: Harmonized Multi-Layer Text-to-Image Generation with Generative Priors

ControlVAR: Exploring Controllable Visual Autoregressive Modeling