CoReDiT reduces self-attention FLOPs in DiTs by up to 55% via linear-time spatial coherence pruning and neighbor-based reconstruction, delivering 1.33x-1.72x speedups with maintained quality.
Token merging for fast sta- ble diffusion
3 Pith papers cite this work. Polarity classification is still indexing.
3
Pith papers citing it
fields
cs.CV 3representative citing papers
DisCa replaces heuristic feature caching with a lightweight learnable neural predictor compatible with distillation, achieving 11.8× acceleration on video diffusion transformers with preserved generation quality.
FastVGGT achieves 4x speedup on VGGT for 1000-image inputs using training-free token merging tailored to 3D architectures while reducing error accumulation.
citing papers explorer
-
CoReDiT: Spatial Coherence-Guided Token Pruning and Reconstruction for Efficient Diffusion Transformers
CoReDiT reduces self-attention FLOPs in DiTs by up to 55% via linear-time spatial coherence pruning and neighbor-based reconstruction, delivering 1.33x-1.72x speedups with maintained quality.
-
DisCa: Accelerating Video Diffusion Transformers with Distillation-Compatible Learnable Feature Caching
DisCa replaces heuristic feature caching with a lightweight learnable neural predictor compatible with distillation, achieving 11.8× acceleration on video diffusion transformers with preserved generation quality.
-
FastVGGT: Training-Free Acceleration of Visual Geometry Transformer
FastVGGT achieves 4x speedup on VGGT for 1000-image inputs using training-free token merging tailored to 3D architectures while reducing error accumulation.