paper-with-me

홈 › Papers

Local Representative Token Guided Merging for Text-to-Image Generation

2025-07-17 · Min-Jeong Lee, Hee-Dong Kim, Seong-Whan Lee arxiv

Stable diffusion is an outstanding image generation model for text-to-image, but its time-consuming generation process remains a challenge due to the quadratic complexity of attention operations. Recent token merging methods improve efficiency by reducing the number of tokens during attention operations, but often overlook the characteristics of attention-based image generation models, limiting their effectiveness. In this paper, we propose local representative token guided merging (ReToM), a novel token merging strategy applicable to any attention mechanism in image generation. To merge tokens based on various contextual information, ReToM defines local boundaries as windows within attention inputs and adjusts window sizes. Furthermore, we introduce a representative token, which represents the most representative token per window by computing similarity at a specific timestep and selecting the token with the highest average similarity. This approach preserves the most salient local features while minimizing computational overhead. Experimental results show that ReToM achieves a 6.2% improvement in FID and higher CLIP scores compared to the baseline, while maintaining comparable inference time. We empirically demonstrate that ReToM is effective in balancing visual quality and computational efficiency.

📄 PDF Abstract BibTeX arXiv:2507.12771

Code (0)

등록된 구현이 없습니다.

Tasks

Text-to-Image GenerationComputational Efficiency

Similar Papers 제목 키워드 기반

StructSAM: Structure- and Spectrum-Preserving Token Merging for Segment Anything Models

2026-03-07 · Duy M. H. Nguyen, Tuan A. Tran, Duong Nguyen, Siwei Xie 외 arxiv

Recent token merging techniques for Vision Transformers (ViTs) provide substantial speedups by reducing the number of tokens processed by self-attention, often without retraining. However, their direct application to the…

FlowMM: Cross-Modal Information Flow Guided KV Cache Merging for Efficient Multimodal Context Inference

2025-10-29 · Kunxi Li, Yufan Xiong, Zhonghua Jiang, Yiyun Zhou 외 arxiv

Traditional KV cache eviction strategies, which discard less critical KV-pairs based on attention scores, often degrade generation quality, causing context loss or hallucinations. Recent efforts shift toward KV merging, …

VisionTrim: Unified Vision Token Compression for Training-Free MLLM Acceleration

2026-01-30 · Hanxun Yu, Wentong Li, Xuan Qu, Song Wang 외 arxiv

Multimodal large language models (MLLMs) suffer from high computational costs due to excessive visual tokens, particularly in high-resolution and video-based scenarios. Existing token reduction methods typically focus on…

ReGATE: Learning Faster and Better with Fewer Tokens in MLLMs

2025-07-29 · Chaoyu Li, Yogesh Kulkarni, Pooyan Fazli arxiv

The computational cost of training multimodal large language models (MLLMs) grows rapidly with the number of processed tokens. Existing efficiency methods mainly target inference via token reduction or merging, offering …

PlaM: Training-Free Plateau-Guided Model Merging for Better Visual Grounding in MLLMs

2026-01-12 · Zijing Wang, Yongkang Liu, Mingyang Wang, Ercong Nie 외 arxiv

Multimodal Large Language Models (MLLMs) rely on strong linguistic reasoning inherited from their base language models. However, multimodal instruction fine-tuning paradoxically degrades this text's reasoning capability,…

Visual Grounding