9 papers
Chamaileon: Cross-Context Binder Design with Contextualized Modeling and Mixed Sampling
Hengyuan Cao, Shizhuo Cheng, Mingxuan Liu +5
The rapid evolution of generative models has unlocked new potentials in protein binder design, a pivotal task in structural biology, by facilitating end-to-end generation via joint…
AAD-1: Asymmetric Adversarial Distillation for One-Step Autoregressive Video Generation
Haobo Li, Yanhong Zeng, Yunhong Lu +6
We present AAD-1, an Asymmetric Adversarial Distillation framework for One-step autoregressive image-to-video generation. State-of-the-art methods adopt adversarial distillation bu…
Offline Preference Optimization for Rectified Flow with Noise-Tracked Pairs
Yunhong Lu, Qichao Wang, Hengyuan Cao +2
Existing preference datasets for text-to-image models typically store only the final winner/loser images. This representation is insufficient for rectified flow (RF) models, whose…
DMGD: Train-Free Dataset Distillation with Semantic-Distribution Matching in Diffusion Models
Qichao Wang, Yunhong Lu, Hengyuan Cao +2
Dataset distillation enables efficient training by distilling the information of large-scale datasets into significantly smaller synthetic datasets. Diffusion based paradigms have…
Spherical Geometry Diffusion: Generating High-quality 3D Face Geometry via Sphere-anchored Representations
Junyi Zhang, Yiming Wang, Yunhong Lu +5
A fundamental challenge in text-to-3D face generation is achieving high-quality geometry. The core difficulty lies in the arbitrary and intricate distribution of vertices in 3D spa…
Reward Forcing: Efficient Streaming Video Generation with Rewarded Distribution Matching Distillation
Yunhong Lu, Yanhong Zeng, Haobo Li +9
Efficient streaming video generation is critical for simulating interactive and dynamic worlds. Existing methods distill few-step video diffusion models with sliding window attenti…
Smoothed Preference Optimization via ReNoise Inversion for Aligning Diffusion Models with Varied Human Preferences
Yunhong Lu, Qichao Wang, Hengyuan Cao +2
Direct Preference Optimization (DPO) aligns text-to-image (T2I) generation models with human preferences using pairwise preference data. Although substantial resources are expended…
Dimension-Reduction Attack! Video Generative Models are Experts on Controllable Image Synthesis
Hengyuan Cao, Yutong Feng, Biao Gong +4
Video generative models can be regarded as world simulators due to their ability to capture dynamic, continuous changes inherent in real-world environments. These models integrate…
InPO: Inversion Preference Optimization with Reparametrized DDIM for Efficient Diffusion Model Alignment
Yunhong Lu, Qichao Wang, Hengyuan Cao +3
Without using explicit reward, direct preference optimization (DPO) employs paired human preference data to fine-tune generative models, a method that has garnered considerable att…