APT: Improving Diffusion Models for High Resolution Image Generation with Adaptive Path Tracing

2025-07-29

AI Summary

Diffusion models face two key challenges in high-resolution image generation: (1) degraded generalization due to fixed-resolution training, and (2) patch-level distribution shift and increased monotonicity induced by patch-wise, training-free inference methods—both compromising detail fidelity and sampling efficiency. To address these issues without additional training, we propose an Adaptive Path Tracing framework that jointly integrates statistical matching and scale-aware scheduling within the latent space, dynamically calibrating patch-wise distributions and optimizing denoising trajectories. Coupled with latent diffusion models and adaptive patch fusion, our method significantly enhances texture sharpness and structural coherence in high-resolution outputs while accelerating sampling. Extensive experiments demonstrate superior performance over existing training-free approaches in standard metrics—including lower FID and LPIPS—achieving a favorable balance between generation quality and inference efficiency.

Technology Category

• Computer Vision: Diffusion Models for Vision
• Planning, Routing, and Scheduling: Planning with Language Models
• Search and Optimization: Sampling/Simulation-based Search

Application Category

• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Latent Diffusion Models (LDMs) are generally trained at fixed resolutions, limiting their capability when scaling up to high-resolution images. While training-based approaches address this limitation by training on high-resolution datasets, they require large amounts of data and considerable computational resources, making them less practical. Consequently, training-free methods, particularly patch-based approaches, have become a popular alternative. These methods divide an image into patches and fuse the denoising paths of each patch, showing strong performance on high-resolution generation. However, we observe two critical issues for patch-based approaches, which we call ``patch-level distribution shift" and ``increased patch monotonicity." To address these issues, we propose Adaptive Path Tracing (APT), a framework that combines Statistical Matching to ensure patch distributions remain consistent in upsampled latents and Scale-aware Scheduling to deal with the patch monotonicity. As a result, APT produces clearer and more refined details in high-resolution images. In addition, APT enables a shortcut denoising process, resulting in faster sampling with minimal quality degradation. Our experimental results confirm that APT produces more detailed outputs with improved inference speed, providing a practical approach to high-resolution image generation.

Problem

• LDMs struggle with high-resolution images due to fixed training resolutions
• Patch-based methods face patch-level distribution shift and monotonicity issues
• APT improves high-resolution generation with adaptive path tracing techniques

Innovation

• Adaptive Path Tracing for high-res images
• Statistical Matching for consistent patch distributions
• Scale-aware Scheduling reduces patch monotonicity