Mean-Shift Distillation for Diffusion Mode Seeking
We present mean-shift distillation, a novel diffusion distillation technique that provides a provably good proxy for the gradient of the diffusion output distribution. This is derived directly from mean-shift mode seeking on the distribution, and we show that its extrema are aligned with the modes. We further derive an efficient product distribution sampling procedure to evaluate the gradient. Our method is formulated as a drop-in replacement for score distillation sampling (SDS), requiring neither model retraining nor extensive modification of the sampling procedure. We show that it exhibits superior mode alignment as well as improved convergence in both synthetic and practical setups, yielding higher-fidelity results when applied to both text-to-image and text-to-3D applications with Stable Diffusion.
Code (0)
등록된 구현이 없습니다.
Tasks
Text to 3DMethods 이 논문이 사용한 방법론
Similar Papers 제목 키워드 기반
Stabilizing, Scaling & Enhancing MeanFlow for Large-scale Diffusion Distillation
Diffusion models exhibit remarkable generative capability, but their high latency limits practical deployment. Many studies have attempted to reduce sampling steps to accelerate inference. Among them, MeanFlow has attrac…
EM Distillation for One-step Diffusion Models
While diffusion models can learn complex distributions, sampling requires a computationally expensive iterative process. Existing distillation methods enable efficient sampling, but have notable limitations, such as perf…
MeanShift++: Extremely Fast Mode-Seeking With Applications to Segmentation and Object Tracking
MeanShift is a popular mode-seeking clustering algorithm used in a wide range of applications in machine learning. However, it is known to be prohibitively slow, with quadratic runtime per iteration. We propose MeanShift…
ClusteringDensity EstimationImage SegmentationObject Tracking+1Adversarial Distribution Matching for Diffusion Distillation Towards Efficient Image and Video Synthesis
Distribution Matching Distillation (DMD) is a promising score distillation technique that compresses pre-trained teacher diffusion models into efficient one-step or multi-step student generators. Nevertheless, its relian…
Improving Knowledge Distillation Under Unknown Covariate Shift Through Confidence-Guided Data Augmentation
Large foundation models trained on extensive datasets demonstrate strong zero-shot capabilities in various domains. Knowledge distillation has become an established tool for transferring knowledge from foundation models …