{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,1,8]],"date-time":"2026-01-08T20:27:56Z","timestamp":1767904076434,"version":"3.49.0"},"reference-count":57,"publisher":"Oxford University Press (OUP)","issue":"1","license":[{"start":{"date-parts":[[2025,11,17]],"date-time":"2025-11-17T00:00:00Z","timestamp":1763337600000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by-nc\/4.0\/"}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"published-print":{"date-parts":[[2026,1,2]]},"abstract":"<jats:title>Abstract<\/jats:title>\n                  <jats:p>This paper presents a novel algorithm that enhances Structure-from-Motion (SfM) pipelines by integrating efficient novel view synthesis (NVS) using 3D Gaussian Splatting (3DGS). Traditional SfM methods often fail in challenging scenarios with repetitive structures, outliers or misaligned images, leading to ghost and doppelganger artifacts. To address this, we propose an NVS-guided evaluation and correction framework that leverages 3DGS-rendered views to detect and rectify misaligned images, improving reconstruction accuracy. Our method not only enhances SfM outputs but also benefits downstream tasks such as dense reconstruction [multi-view stereo (MVS)], NVS, and visual localization. Experiments on public datasets and our real-world indoor dataset show that state-of-the-art baselines fail in several ambiguous scenes where our approach succeeds. For example, on the Cambridge Landmarks dataset (Shop Facade), our method reduces localization error from 5.7 to 5.4\u00a0cm and improves SSIM from 0.735 to 0.787. These results confirm the robustness and broad applicability of our framework, demonstrating that while 3DGS is an effective tool, the core contribution lies in the proposed SfM re-alignment strategy.<\/jats:p>","DOI":"10.1093\/jcde\/qwaf125","type":"journal-article","created":{"date-parts":[[2025,11,14]],"date-time":"2025-11-14T13:03:24Z","timestamp":1763125404000},"page":"303-323","source":"Crossref","is-referenced-by-count":0,"title":["Enhancing Structure-from-Motion: Re-aligning images with 3D Gaussian splatting"],"prefix":"10.1093","volume":"13","author":[{"ORCID":"https:\/\/orcid.org\/0000-0002-9436-3448","authenticated-orcid":false,"given":"Daewoon","family":"Kim","sequence":"first","affiliation":[{"name":"Tech. Innovation Group , KT, 151 Taebong-ro, Seocho-gu, Seoul 06763 ,","place":["Republic of Korea"]}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Jinwook","family":"Park","sequence":"additional","affiliation":[{"name":"Tech. Innovation Group , KT, 151 Taebong-ro, Seocho-gu, Seoul 06763 ,","place":["Republic of Korea"]}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0009-0001-4938-0038","authenticated-orcid":false,"given":"I-gil","family":"Kim","sequence":"additional","affiliation":[{"name":"Tech. Innovation Group , KT, 151 Taebong-ro, Seocho-gu, Seoul 06763 ,","place":["Republic of Korea"]}],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"286","published-online":{"date-parts":[[2025,11,17]]},"reference":[{"key":"2026010810574400500_bib1","doi-asserted-by":"publisher","first-page":"105","DOI":"10.1145\/2001269.2001293","article-title":"Building rome in a day","volume":"54","author":"Agarwal","year":"2011","journal-title":"Commun. ACM"},{"key":"2026010810574400500_bib2","first-page":"5297","article-title":"Netvlad:cnn architecture for weakly supervised place recognition","volume-title":"Proc. IEEE Conf. Comput. Vis. Pattern Recognit. (CVPR)","author":"Arandjelovi\u0107","year":"2016"},{"key":"2026010810574400500_bib3","first-page":"751","article-title":"Relocnet: Continuous metric learning relocalisation using neural nets","volume-title":"Proc. Eur. Conf. Comput. Vis. (ECCV)","author":"Balntas","year":"2018"},{"key":"2026010810574400500_bib4","first-page":"1","article-title":"Matching 2d images in 3d: Metric relative pose from metric correspondences","volume-title":"Proc. IEEE\/CVF Conf. Comput. Vis. Pattern Recognit. (CVPR)","author":"Barroso-Laguna","year":"2024"},{"key":"2026010810574400500_bib5","first-page":"404","article-title":"Surf: Speeded up robust features","volume-title":"Proc. Eur. Conf. Comput. Vis. (ECCV)","author":"Bay","year":"2006"},{"key":"2026010810574400500_bib6","first-page":"4654","article-title":"Learning less is more: 6d camera localization via 3d surface regression","volume-title":"Proc. IEEE Conf. Comput. Vis. Pattern Recognit. (CVPR)","author":"Brachmann","year":"2018"},{"key":"2026010810574400500_bib7","first-page":"3070","article-title":"Visual camera re-localization from rgb and rgb-d images using dsac","volume":"43","author":"Brachmann","year":"2021","journal-title":"IEEE Trans. Pattern Anal. Mach. Intell."},{"key":"2026010810574400500_bib8","first-page":"6684","article-title":"Dsac: Differentiable ransac for camera localization","volume-title":"Proc. IEEE Conf. Comput. Vis. Pattern Recognit. (CVPR)","author":"Brachmann","year":"2017"},{"key":"2026010810574400500_bib9","first-page":"2616","article-title":"Geometry-aware learning of maps for camera localization","volume-title":"Proc. IEEE Conf. Comput. Vis. Pattern Recognit. (CVPR)","author":"Brachmann","year":"2017"},{"key":"2026010810574400500_bib10","first-page":"20796","article-title":"Accelerated coordinate encoding: Learning to relocalize in minutes using rgb and poses","volume-title":"Proc. IEEE\/CVF Conf. Comput. Vis. Pattern Recognit. (CVPR)","author":"Brachmann","year":"2023"},{"key":"2026010810574400500_bib11","first-page":"1","article-title":"Doppelgangers: Learning to disambiguate images of similar structures","volume-title":"Proc. IEEE\/CVF Int. Conf. Comput. Vis. (ICCV)","author":"Cai","year":"2023"},{"key":"2026010810574400500_bib12","article-title":"A survey on 3D Gaussian splatting","author":"Chen","year":"2025"},{"key":"2026010810574400500_bib13","article-title":"Gaussianpro: 3D Gaussian splatting with progressive propagation","author":"Cheng","year":"2024"},{"key":"2026010810574400500_bib14","first-page":"864","article-title":"Global structure-from-motion by similarity averaging","volume-title":"Proc. IEEE Int. Conf. Comput. Vis. (ICCV)","author":"Cui","year":"2015"},{"key":"2026010810574400500_bib15","first-page":"337","article-title":"Superpoint: Self-supervised interest point detection and description","volume-title":"Proc. IEEE Conf. Comput. Vis. Pattern Recognit. Workshops (CVPRW)","author":"DeTone","year":"2018"},{"key":"2026010810574400500_bib16","first-page":"2871","article-title":"Camnet: Coarse-to-fine retrieval for camera re-localization","volume-title":"Proc. IEEE\/CVF Int. Conf. Comput. Vis. (ICCV)","author":"Ding","year":"2019"},{"key":"2026010810574400500_bib17","first-page":"3859","article-title":"A critique of structure-from-motion algorithms","volume-title":"Proc. IEEE Int. Conf. Robot. Autom. (ICRA)","author":"Eade","year":"2013"},{"key":"2026010810574400500_bib18","first-page":"20796","article-title":"Colmap-free 3D Gaussian splatting","volume-title":"Proc. IEEE\/CVF Conf. Comput. Vis. Pattern Recognit. (CVPR)","author":"Fu","year":"2024"},{"key":"2026010810574400500_bib19","first-page":"1","volume-title":"Multiple View Geometry in Computer Vision","author":"Hartley","year":"2003","edition":"2nd edn"},{"key":"2026010810574400500_bib20","first-page":"788","article-title":"Correcting for duplicate scene structure in sparse 3d reconstruction","volume-title":"Proc. Eur. Conf. Comput. Vis. (ECCV)","author":"Heinly","year":"2014"},{"key":"2026010810574400500_bib21","first-page":"4762","article-title":"Modelling uncertainty in deep learning for camera relocalization","volume-title":"Proc. IEEE Int. Conf. Robot. Autom. (ICRA)","author":"Kendall","year":"2016"},{"key":"2026010810574400500_bib22","first-page":"2938","article-title":"Posenet: A convolutional network for real-time 6-dof camera relocalization","volume-title":"Proc. IEEE Int. Conf. Comput. Vis. (ICCV)","author":"Kendall","year":"2015"},{"key":"2026010810574400500_bib23","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1145\/3592433","article-title":"3D Gaussian splatting for real-time radiance field rendering","volume":"42","author":"Kerbl","year":"2023","journal-title":"ACM Trans. Graph. (SIGGRAPH)"},{"key":"2026010810574400500_bib24","doi-asserted-by":"publisher","first-page":"1482","DOI":"10.1093\/jcde\/qwac066","article-title":"Camera localization with siamese neural networks using iterative relative pose estimation","volume":"9","author":"Kim","year":"2022","journal-title":"Journal of Computational Design and Engineering"},{"key":"2026010810574400500_bib25","doi-asserted-by":"publisher","first-page":"1047","DOI":"10.1093\/jcde\/qwad035","article-title":"Augmented virtual reality and 360 spatial visualization for supporting user-engaged design","volume":"10","author":"Lee","year":"2023","journal-title":"Journal of Computational Design and Engineering"},{"key":"2026010810574400500_bib26","first-page":"5742","article-title":"Barf: Bundle-adjusting neural radiance fields","volume-title":"Proc. IEEE\/CVF Int. Conf. Comput. Vis. (ICCV)","author":"Lin","year":"2021"},{"key":"2026010810574400500_bib27","first-page":"5987","article-title":"Pixel-perfect structure-from-motion with featuremetric refinement","volume-title":"Proc. IEEE\/CVF Int. Conf. Comput. Vis. (ICCV)","author":"Lindenberger","year":"2021"},{"key":"2026010810574400500_bib28","first-page":"1","article-title":"Lightglue: Local feature matching at light speed","volume-title":"Proc. IEEE\/CVF Int. Conf. Comput. Vis. (ICCV)","author":"Lindenberger","year":"2023"},{"key":"2026010810574400500_bib29","doi-asserted-by":"publisher","first-page":"91","DOI":"10.1023\/B:VISI.0000029664.99615.94","article-title":"Distinctive image features from scale-invariant keypoints","volume":"60","author":"Lowe","year":"2004","journal-title":"Int. J. Comput. Vis."},{"key":"2026010810574400500_bib30","doi-asserted-by":"publisher","first-page":"052009","DOI":"10.1088\/1742-6596\/1087\/5\/052009","article-title":"A review of solutions for perspective-n-point problem in camera pose estimation","volume":"1087","author":"Lu","year":"2018","journal-title":"J. Phys.: Conf. Ser."},{"key":"2026010810574400500_bib31","first-page":"405","article-title":"Nerf: Representing scenes as neural radiance fields for view synthesis","volume-title":"Proc. Eur. Conf. Comput. Vis. (ECCV)","author":"Mildenhall","year":"2020"},{"key":"2026010810574400500_bib32","first-page":"3248","article-title":"Global fusion of relative motions for robust, accurate and scalable structure from motion","volume-title":"Proc. IEEE Int. Conf. Comput. Vis. (ICCV)","author":"Moulon","year":"2013"},{"key":"2026010810574400500_bib33","first-page":"60","article-title":"Openmvg: Open multiple view geometry","volume-title":"Reproducible Research in Pattern Recognition, vol. 10214 of Lecture Notes in Computer Science","author":"Moulon","year":"2016"},{"key":"2026010810574400500_bib34","doi-asserted-by":"crossref","first-page":"102:1","DOI":"10.1145\/3528223.3530127","article-title":"Instant neural graphics primitives with a multiresolution hash encoding","volume":"41","author":"M\u00fcller","year":"2022","journal-title":"ACM Trans. Graph. (SIGGRAPH)"},{"key":"2026010810574400500_bib35","first-page":"1","article-title":"Focustune: Tuning visual localization through focus-guided sampling","volume-title":"Proc. IEEE\/CVF Winter Conf. Appl. Comput. Vis. (WACV)","author":"Nguyen","year":"2024"},{"key":"2026010810574400500_bib36","first-page":"8748","article-title":"Learning transferable visual models from natural language supervision","volume-title":"Proc. Int. Conf. Mach. Learn. (ICML)","author":"Radford","year":"2021"},{"key":"2026010810574400500_bib37","first-page":"3137","article-title":"Structure from motion for scenes with large duplicate structures","volume-title":"Proc. IEEE Conf. Comput. Vis. Pattern Recognit. (CVPR)","author":"Roberts","year":"2011"},{"key":"2026010810574400500_bib38","first-page":"12708","article-title":"From coarse to fine: Robust hierarchical localization at large scale","volume-title":"Proc. IEEE\/CVF Conf. Comput. Vis. Pattern Recognit. (CVPR)","author":"Sarlin","year":"2019"},{"key":"2026010810574400500_bib39","first-page":"4938","article-title":"Superglue: Learning feature matching with graph neural networks","volume-title":"Proc. IEEE\/CVF Conf. Comput. Vis. Pattern Recognit. (CVPR)","author":"Sarlin","year":"2020"},{"key":"2026010810574400500_bib40","first-page":"4104","article-title":"Structure-from-motion revisited","volume-title":"Proc. IEEE Conf. Comput. Vis. Pattern Recognit. (CVPR)","author":"Schonberger","year":"2016"},{"key":"2026010810574400500_bib41","doi-asserted-by":"publisher","first-page":"835","DOI":"10.1145\/1141911.1141964","article-title":"Photo tourism: Exploring photo collections in 3d","volume":"25","author":"Snavely","year":"2006","journal-title":"ACM Trans. Graph. (SIGGRAPH)"},{"key":"2026010810574400500_bib42","doi-asserted-by":"publisher","first-page":"1097","DOI":"10.1093\/jcde\/qwac046","article-title":"Learning-based essential matrix estimation for visual localization","volume":"9","author":"Son","year":"2022","journal-title":"Journal of Computational Design and Engineering"},{"key":"2026010810574400500_bib43","first-page":"8922","article-title":"Loftr: Detector-free local feature matching with transformers","volume-title":"Proc. IEEE\/CVF Conf. Comput. Vis. Pattern Recognit. (CVPR)","author":"Sun","year":"2021"},{"key":"2026010810574400500_bib44","first-page":"298","article-title":"Bundle adjustment\u2013a modern synthesis","volume-title":"Proc. Int. Workshop Vis. Algorithms","author":"Triggs","year":"1999"},{"key":"2026010810574400500_bib45","first-page":"14254","article-title":"Disk: Learning local features with policy gradient","volume-title":"Proc. Adv. Neural Inf. Process. Syst. (NeurIPS)","author":"Tyszkiewicz","year":"2020"},{"key":"2026010810574400500_bib46","first-page":"1","article-title":"Glace: Global local accelerated coordinate encoding","volume-title":"Proc. IEEE\/CVF Conf. Comput. Vis. Pattern Recognit. (CVPR)","author":"Wang","year":"2024"},{"key":"2026010810574400500_bib47","first-page":"1","article-title":"Vggsfm: Visual geometry grounded deep structure from motion","volume-title":"Proc. IEEE\/CVF Conf. Comput. Vis. Pattern Recognit. (CVPR)","author":"Wang","year":"2024"},{"key":"2026010810574400500_bib48","first-page":"5294","article-title":"Vggt: Visual geometry grounded transformer","volume-title":"Proceedings of the Computer Vision and Pattern Recognition Conference","author":"Wang","year":"2025"},{"key":"2026010810574400500_bib49","doi-asserted-by":"publisher","first-page":"600","DOI":"10.1109\/TIP.2003.819861","article-title":"Image quality assessment: From error visibility to structural similarity","volume":"13","author":"Wang","year":"2004","journal-title":"IEEE Trans. Image Process."},{"key":"2026010810574400500_bib50","first-page":"513","article-title":"Network principles for sfm: Disambiguating repeated structures with local context","volume-title":"Proc. IEEE Int. Conf. Comput. Vis. (ICCV)","author":"Wilson","year":"2013"},{"key":"2026010810574400500_bib51","first-page":"255","article-title":"Robust global translations with 1dsfm","volume-title":"Proc. Eur. Conf. Comput. Vis. (ECCV)","author":"Wilson","year":"2014"},{"key":"2026010810574400500_bib52","article-title":"Visualsfm: A visual structure from motion system","author":"Wu","year":"2011"},{"key":"2026010810574400500_bib53","first-page":"1","article-title":"Doppelgangers++: Improved visual disambiguation with geometric 3d features","volume-title":"Proc. IEEE\/CVF Conf. Comput. Vis. Pattern Recognit. (CVPR)","author":"Xiangli","year":"2024"},{"key":"2026010810574400500_bib54","article-title":"An efficient scene coordinate encoding and relocalization method","author":"Xu","year":"2024"},{"key":"2026010810574400500_bib55","doi-asserted-by":"publisher","first-page":"70","DOI":"10.1093\/jcde\/qwaf116","article-title":"Dvs-3d: Diffusion-based novel view synthesis and 3d object reconstruction from a single image","volume":"12","author":"Xu","year":"2025","journal-title":"Journal of Computational Design and Engineering"},{"key":"2026010810574400500_bib56","first-page":"3836","article-title":"Distinguishing the indistinguishable: Exploring structural ambiguities via geodesic context","volume-title":"Proc. IEEE Conf. Comput. Vis. Pattern Recognit. (CVPR)","author":"Yan","year":"2017"},{"key":"2026010810574400500_bib57","first-page":"341","article-title":"Gaussian in the wild: 3D Gaussian splatting for unconstrained image collections","volume-title":"European Conference on Computer Vision","author":"Zhang","year":"2024"}],"container-title":["Journal of Computational Design and Engineering"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/academic.oup.com\/jcde\/advance-article-pdf\/doi\/10.1093\/jcde\/qwaf125\/65356700\/qwaf125.pdf","content-type":"application\/pdf","content-version":"am","intended-application":"syndication"},{"URL":"https:\/\/academic.oup.com\/jcde\/article-pdf\/13\/1\/303\/65356700\/qwaf125.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"syndication"},{"URL":"https:\/\/academic.oup.com\/jcde\/article-pdf\/13\/1\/303\/65356700\/qwaf125.pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,1,8]],"date-time":"2026-01-08T15:57:59Z","timestamp":1767887879000},"score":1,"resource":{"primary":{"URL":"https:\/\/academic.oup.com\/jcde\/article\/13\/1\/303\/8325207"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,11,17]]},"references-count":57,"journal-issue":{"issue":"1","published-print":{"date-parts":[[2026,1,2]]}},"URL":"https:\/\/doi.org\/10.1093\/jcde\/qwaf125","relation":{},"ISSN":["2288-5048"],"issn-type":[{"value":"2288-5048","type":"electronic"}],"subject":[],"published-other":{"date-parts":[[2026,1]]},"published":{"date-parts":[[2025,11,17]]}}}