9 papers
Frequency-Step Cache: Separating Synchronization and Repair for Efficient Autoregressive Video Diffusion
Kerry Xu, Param Misra, Zachary Stenglein +3
Autoregressive (AR) video generation enables long-horizon synthesis but requires repeated forward passes through large diffusion transformers. Training-free caching reduces this co…
Splaxel: Efficient Distributed Training of 3D Gaussian Splatting for Large-scale Scene Reconstruction via Pixel-level Communication
Wenqi Jia, Zhewen Hu, Ying Huang +10
3D Gaussian Splatting (3DGS) enables high-fidelity and real-time 3D scene reconstruction, but scaling training to large-scale scenes requires optimizing hundreds of millions of Gau…
GENSERVE: Efficient Co-Serving of Heterogeneous Diffusion Model Workloads
Fanjiang Ye, Zhangke Li, Xinrui Zhong +10
Diffusion models have emerged as the prevailing approach for text-to-image (T2I) and text-to-video (T2V) generation, yet production platforms must increasingly serve both modalitie…
HII-DPO: Eliminate Hallucination via Accurate Hallucination-Inducing Counterfactual Images
Yilin Yang, Zhenghui Guo, Yuke Wang +3
Large Vision-Language Models (VLMs) have achieved remarkable success across diverse multimodal tasks but remain vulnerable to hallucinations rooted in inherent language bias. Despi…
SDiT: Semantic Region-Adaptive for Diffusion Transformers
Bowen Lin, Fanjiang Ye, Yihua Liu +7
Diffusion Transformers (DiTs) achieve state-of-the-art performance in text-to-image synthesis but remain computationally expensive due to the iterative nature of denoising and the…
An Efficient and Adaptive Watermark Detection System with Tile-based Error Correction
Xinrui Zhong, Xinze Feng, Jingwei Zuo +6
Efficient and reliable detection of generated images is critical for the responsible deployment of generative models. Existing approaches primarily focus on improving detection acc…
SuperGen: An Efficient Ultra-high-resolution Video Generation System with Sketching and Tiling
Fanjiang Ye, Zepeng Zhao, Yi Mu +11
Diffusion models have recently achieved remarkable success in generative tasks (e.g., image and video generation), and the demand for high-quality content (e.g., 2K/4K videos) is r…
Utilizing Sequential Information of General Lab-test Results and Diagnoses History for Differential Diagnosis of Dementia
Yizong Xing, Dhita Putri Pratama, Yuke Wang +2
Early diagnosis of Alzheimer's Disease (AD) faces multiple data-related challenges, including high variability in patient data, limited access to specialized diagnostic tests, and…
PATCHEDSERVE: A Patch Management Framework for SLO-Optimized Hybrid Resolution Diffusion Serving
Desen Sun, Zepeng Zhao, Yuke Wang
The Text-to-Image (T2I) diffusion model has emerged as one of the most widely adopted generative models. However, serving diffusion models at the granularity of entire images intro…