17 papers
TTS-Q: Speaker-Aware Post-Training Quantization of Text-to-Speech Diffusion Transformers
Xiaoqin Feng, Yuqi Li, Weilun Feng +9
Zero-shot text-to-speech (TTS) can clone a voice from seconds of audio. Yet the quality comes from billion-parameter diffusion transformers whose size and multi-step sampling put t…
SceneScaffold: Active Scene-State Construction for Unified 3D Scene Understanding
Xiangqi Li, Libo Huang, Jiarui Zhao +4
Recent 3D large multimodal models (3D-LMMs) rely on a visual bottleneck to compress complex 3D scene evidence into a limited number of visual tokens compatible with large language…
Towards Compact Unified Multimodal Tracking: Synergizing Knowledge Distillation with Structural Pruning
Yuqi Li, Yuedong Tan, Huiran Duan +7
Unified multimodal object tracking has achieved remarkable robustness by leveraging complementary sensor data (e.g., RGB, Thermal, Depth), yet the heavy computational burden of sta…
Multi-Resolution Flow Matching: Training-Free Diffusion Acceleration via Staged Sampling
Xingyu Zheng, Xianglong Liu, Yifu Ding +4
Hardware-agnostic strategies for accelerating text-to-image diffusion, such as timestep distillation and feature caching, can reduce inference time without custom kernels or system…
Not All Tasks Quantize Equally: Fisher-Guided Quantization for Visual Geometry Transformer
Yipu Zhang, Jintao Cheng, Weilun Feng +5
Feed-forward 3D reconstruction models, represented by Visual Geometry Grounded Transformer (VGGT), jointly predict multiple visual geometry tasks such as depth estimation, camera p…
Echo-Forcing: A Scene Memory Framework for Interactive Long Video Generation
Mingqiang Wu, Weilun Feng, Zhefeng Zhang +8
Autoregressive video diffusion models enable open-ended generation through local attention and KV caching. However, existing training-free long-video optimization methods mainly fo…
WorldCache: Accelerating World Models for Free via Heterogeneous Token Caching
Weilun Feng, Guoxin Fan, Haotong Qin +10
Diffusion-based world models have shown strong potential for unified world simulation, but the iterative denoising remains too costly for interactive use and long-horizon rollouts.…
Teacher-Guided Student Self-Knowledge Distillation Using Diffusion Model
Yu Wang, Chuanguang Yang, Zhulin An +6
Existing Knowledge Distillation (KD) methods often align feature information between teacher and student by exploring meaningful feature processing and loss functions. However, due…
Fast-SAM3D: 3Dfy Anything in Images but Faster
Weilun Feng, Mingqiang Wu, Zhiliang Chen +10
SAM3D enables scalable, open-world 3D reconstruction from complex scenes, yet its deployment is hindered by prohibitive inference latency. In this work, we conduct the \textbf{firs…
QuantSparse: Comprehensively Compressing Video Diffusion Transformer with Model Quantization and Attention Sparsification
Weilun Feng, Chuanguang Yang, Haotong Qin +8
Diffusion transformers exhibit remarkable video generation capability, yet their prohibitive computational and memory costs hinder practical deployment. Model quantization and atte…
Quantized Visual Geometry Grounded Transformer
Weilun Feng, Haotong Qin, Mingqiang Wu +8
Learning-based 3D reconstruction models, represented by Visual Geometry Grounded Transformers (VGGTs), have made remarkable progress with the use of large-scale transformers. Their…
SQ-VDiT: Accurate Quantized Video Diffusion Transformer with Salient Data and Sparse Token Distillation
Weilun Feng, Haotong Qin, Chuanguang Yang +7
Diffusion transformers have emerged as the mainstream paradigm for video generation models. However, the use of up to billions of parameters incurs significant computational costs.…
MPQ-DMv2: Flexible Residual Mixed Precision Quantization for Low-Bit Diffusion Models with Temporal Distillation
Weilun Feng, Chuanguang Yang, Haotong Qin +10
Diffusion models have demonstrated remarkable performance on vision generation tasks. However, the high computational complexity hinders its wide application on edge devices. Quant…
Q-VDiT: Towards Accurate Quantization and Distillation of Video-Generation Diffusion Transformers
Weilun Feng, Chuanguang Yang, Haotong Qin +8
Diffusion transformers (DiT) have demonstrated exceptional performance in video generation. However, their large number of parameters and high computational complexity limit their…
Multi-party Collaborative Attention Control for Image Customization
Han Yang, Chuanguang Yang, Qiuli Wang +4
The rapid advancement of diffusion models has increased the need for customized image generation. However, current customization methods face several limitations: 1) typically acce…
MPQ-DM: Mixed Precision Quantization for Extremely Low Bit Diffusion Models
Weilun Feng, Haotong Qin, Chuanguang Yang +7
Diffusion models have received wide attention in generation tasks. However, the expensive computation cost prevents the application of diffusion models in resource-constrained scen…
Relational Diffusion Distillation for Efficient Image Generation
Weilun Feng, Chuanguang Yang, Zhulin An +4
Although the diffusion model has achieved remarkable performance in the field of image generation, its high inference delay hinders its wide application in edge devices with scarce…