DMAligner: Enhancing Image Alignment via Diffusion Model Based View Synthesis
Xinglong Luo, Ao Luo, Zhengning Wang, Yueqi Yang, Chaoyu Feng, Lei Lei, Bing Zeng, Shuaicheng Liu
Abstract
Image alignment is a fundamental task in computer vision with broad applications. Existing methods predominantly employ optical flow-based image warping. However, this technique is susceptible to common challenges such as occlusions and illumination variations, leading to degraded alignment visual quality and compromised accuracy in downstream tasks. In this paper, we present DMAligner, a diffusion-based framework for image alignment through alignment-oriented view synthesis. DMAligner is crafted to tackle the challenges in image alignment from a new perspective, employing a generation-based solution that showcases strong capabilities and avoids the problems associated with flow-based image warping. Specifically, we propose a Dynamics-aware Diffusion Training approach for learning conditional image generation, synthesizing a novel view for image alignment. This incorporates a Dynamics-aware Mask Producing (DMP) module to adaptively distinguish dynamic foreground regions from static backgrounds, enabling the diffusion model to more effectively handle challenges that classical methods struggle to solve. Furthermore, we develop the Dynamic Scene Image Alignment (DSIA) dataset using Blender, which includes 1,033 indoor and outdoor scenes with over 30K image pairs tailored for image alignment. Extensive experimental results demonstrate the superiority of the proposed approach on DSIA benchmarks, as well as on a series of widely-used video datasets for qualitative comparisons. Our code is available at https://github.com/ boomluo02/DMAligner.
Ask Lune
Know this paper in minutes, not an afternoon.
Lune can read every page of it and the top-tier papers that cite it, and gives you its claims, its evidence and its limits, each one cited to the paper it comes from. No skimming, no AI slop.
1 top-tier paper has cited it since. Know what they found before your reviewers do.
Lune can draw on
the full paper · the top-tier papers that cite it · related work
Free to start. No credit card required.
Lineage
- DreamSim: Learning New Dimensions of Human Visual Similarity using Synthetic Data cites High-Resolution Image Synthesis with Latent Diffusion Models
- High-Resolution Image Synthesis with Latent Diffusion Models cites Denoising Diffusion Probabilistic Models
- High-Resolution Image Synthesis with Latent Diffusion Models cites Denoising Diffusion Implicit Models
- Efficient Hybrid SE(3)-Equivariant Visuomotor Flow Policy via Spherical Harmonics for Robot Manipulation cites Denoising Diffusion Probabilistic Models
- Denoising Diffusion Implicit Models cites Denoising Diffusion Probabilistic Models
Cited by top-tier papers1
Builds on35
- Denoising Diffusion Probabilistic ModelsJonathan Ho, Ajay Jain, Pieter AbbeelNeurIPS 2020 · 35,902 citations
- High-Resolution Image Synthesis with Latent Diffusion ModelsRobin Rombach, Andreas Blattmann, Dominik Lorenz, Patrick Esser et al.CVPR 2022 · 13,123 citations
- Denoising Diffusion Implicit ModelsJiaming Song, Chenlin Meng, Stefano ErmonICLR 2021 · 11,743 citations
- DreamSim: Learning New Dimensions of Human Visual Similarity using Synthetic DataStephanie Fu, Netanel Tamir, Shobhita Sundaram, Lucy Chai et al.NeurIPS 2023 · 413 citations
- GMFlow: Learning Optical Flow via Global MatchingHaofei Xu, Jing Zhang, Jianfei Cai, Hamid Rezatofighi et al.CVPR 2022 · 353 citations
Related papers
- Vivid-ZOO: Multi-View Video Generation with Diffusion ModelBing Li, Cheng Zheng, Wenxuan Zhu, Jinjie Mai et al.NeurIPS 2024 · 48 citations
- MoMaps: Semantics-Aware Scene Motion Generation with Motion MapsJiahui Lei, Kyle Genova, George Kopanas, Noah Snavely et al.ICCV 2025 · 1 citation
- Generative Video MattingYongtao Ge, Kangyang Xie, Guangkai Xu, Li Ke et al.SIGGRAPH 2025 · 1 citation
- ViDAR: Video Diffusion-Aware 4D Reconstruction From Monocular InputsMichal Nazarczuk, Sibi Catley-Chandar, Thomas Tanay, Zhensong Zhang et al.NeurIPS 2025 · 5 citations
- DeOcc-1-to-3: 3D De-Occlusion from a Single Image via Self-Supervised Multi-View DiffusionYansong Qu, Shaohui Dai, Xinyang Li, Yuze Wang et al.AAAI 2026
