Saved in:
Bibliographic Details
Main Authors: Wang, Chenglin, Zhou, Yucheng, Chen, Shawn, Wang, Tao, Zhang, Kai
Format: Preprint
Published: 2026
Subjects:
Online Access:https://arxiv.org/abs/2603.13450
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866913020601434112
author Wang, Chenglin
Zhou, Yucheng
Chen, Shawn
Wang, Tao
Zhang, Kai
author_facet Wang, Chenglin
Zhou, Yucheng
Chen, Shawn
Wang, Tao
Zhang, Kai
contents Discrete Diffusion Language Models have emerged as a compelling paradigm for unified multimodal generation, yet their deployment is hindered by high inference latency arising from iterative decoding. Existing acceleration strategies often require expensive re-training or fail to leverage the 2D spatial redundancy inherent in visual data. To address this, we propose Locality-Aware Dynamic Rescue (LADR), a training-free method that expedites inference by exploiting the spatial Markov property of images. LADR prioritizes the recovery of tokens at the ''generation frontier'', regions spatially adjacent to observed pixels, thereby maximizing information gain. Specifically, our method integrates morphological neighbor identification to locate candidate tokens, employs a risk-bounded filtering mechanism to prevent error propagation, and utilizes manifold-consistent inverse scheduling to align the diffusion trajectory with the accelerated mask density. Extensive experiments on four text-to-image generation benchmarks demonstrate that our LADR achieves an approximate 4 x speedup over standard baselines. Remarkably, it maintains or even enhances generative fidelity, particularly in spatial reasoning tasks, offering a state-of-the-art trade-off between efficiency and quality.
format Preprint
id arxiv_https___arxiv_org_abs_2603_13450
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle LADR: Locality-Aware Dynamic Rescue for Efficient Text-to-Image Generation with Diffusion Large Language Models
Wang, Chenglin
Zhou, Yucheng
Chen, Shawn
Wang, Tao
Zhang, Kai
Computer Vision and Pattern Recognition
Computation and Language
Discrete Diffusion Language Models have emerged as a compelling paradigm for unified multimodal generation, yet their deployment is hindered by high inference latency arising from iterative decoding. Existing acceleration strategies often require expensive re-training or fail to leverage the 2D spatial redundancy inherent in visual data. To address this, we propose Locality-Aware Dynamic Rescue (LADR), a training-free method that expedites inference by exploiting the spatial Markov property of images. LADR prioritizes the recovery of tokens at the ''generation frontier'', regions spatially adjacent to observed pixels, thereby maximizing information gain. Specifically, our method integrates morphological neighbor identification to locate candidate tokens, employs a risk-bounded filtering mechanism to prevent error propagation, and utilizes manifold-consistent inverse scheduling to align the diffusion trajectory with the accelerated mask density. Extensive experiments on four text-to-image generation benchmarks demonstrate that our LADR achieves an approximate 4 x speedup over standard baselines. Remarkably, it maintains or even enhances generative fidelity, particularly in spatial reasoning tasks, offering a state-of-the-art trade-off between efficiency and quality.
title LADR: Locality-Aware Dynamic Rescue for Efficient Text-to-Image Generation with Diffusion Large Language Models
topic Computer Vision and Pattern Recognition
Computation and Language
url https://arxiv.org/abs/2603.13450