3DOMFBK

dont miss the deadline for our joint #PhD with @UniPadova on "Multi-modal data fusion for accurate detection of forest parameters". How to apply: tesaf.unipd.it/en/research/ph… @FBK_research #forestry #3D #mapping

also this year, LOW-COST 3D & OPTICAL 3D METROLOGY workshops will team up for a joint event, in Braunschweig (Germany), on 18-20 November 2026. Visit both websites for more info: - lc3d.fbk.eu - o3dm.fbk.eu @FBK_research #photogrammetry #metrology

dont miss our #CVPR2026 workshop on Image Matching on June 4, 2026 image-matching-workshop.github.io and its open competition! @FBK_research @CVPR

Today we launch OpenPizza3D! Our new #3D reconstruction tool optimized for the non-rigid motion of melting mozzarella, tomato sauce volume computation & crust-to-topping ratio with sub-mm precision. ROS2 (Radically Overcooked Slices) plugin coming soon! #AprilFools @FBK_research

We review & synthesize recent advances across opportunistic, infrastructure-based and environment-based #Positioning, #Navigation & #Timing (#PNT) methods, highlighting application scenarios, achievable accuracy & limitations ingentaconnect.com/content/asprs/… #GNSS @OhioState @FBK_research

"3DSceneEditor: Controllable 3D Scene Editing with Gaussian Splatting" ziyangyan.github.io/3DSceneEditor/ a 3D-based pipeline for real-time editing of intricate 3D scenes using #GaussianSplatting. It enables direct manipulation of Gaussians based on input prompts. #WACV2026 @FBK_research

A semi-automatic vectorization of #aerial #photo indexes (APIs) - often called finding aids for aerial reconnaissance sorties - based on computer vision techniques, deep learning and computational geometry ingentaconnect.com/content/asprs/… @FBK_research @CNRsocial_ @MiC_Italia

Accurate #forest inventories are crucial for #ecological research & resource management. We introduce a novel workflow to fuse #pointclouds from TLS, MLS & ULS #LiDAR to improve the characterization of forest structure: ingentaconnect.com/content/asprs/… Pre-print: drive.google.com/file/d/1o_CwuS…

We present a unique #benchmark with dense (1000+ pts/m2) and sparse (35 pts/m2) #multispectral #ALS data for more than 6000 trees for #deeplearning methods for #tree #species #classification: sciencedirect.com/science/articl… @FBK_research @geodepartment @fgi_nls @nibio @SmartForest_SFI

A novel annotation-free open-vocabulary method for #3D #classification of large-scale #aerial #pointclouds: mdpi.com/2072-4292/18/4… - Prior class definitions or annotated data are no longer necessary - Metrics are almost on par with traditional #deeplearning methods @FBK_research

#CVPR2026 workshop "Spatial Intelligence for Cultural Heritage” sint4ch.fbk.eu - Spatial computing & #AI for cultural heritage - #3D modelling, digital twins and #VR - AI for object recognition & analysis in archaeological contexts Submissions deadline: March 1st, 2026

⏳-25! ⌛️ 🖼️Image Matching Workshop 🎯 @#CVPR2026 🏫 👉image-matching-workshop.github.io👈 🏅 Image Matching Challenge 🏆 @kaggle 🖥️ 👉kaggle.com/competitions/i…👈 🧑‍💻@amy_tabb @ducha_aiki @3DOMFBK @kwangmoo_yi @lcmorelli3

Country Overshoot Days 2026 - Earth Overshoot Day overshoot.footprintnetwork.org/newsroom/count…

we evaluate 3D #GaussianSplatting techniques applied to urban scale #3D #reconstruction, analyzing their performance in terms of scalability, geometric accuracy, rendering quality & computational efficiency isprs-archives.copernicus.org/articles/XLVII… #AI #mapping #photogrammetry @FBK_research

-> we release a repo of multi-modal, multi-sensor & multi-platform images -> we present a deep learning-based multi-modal image matching strategy to fuse aerial and terrestrial data Paper: isprs-archives.copernicus.org/articles/XLVII… Repo: github.com/3DOM-FBK/Therm… #heritage #3D @FBK_research

Monocular Depth Estimation (#MDE) has become a valuable tool in #3D #reconstruction. We evaluate MDE algorithms on industrial objects featuring reflections, transparencies or lack of texture -> isprs-archives.copernicus.org/articles/XLVII… #industry @FBK_research #photogrammetry

we analyse the transition from traditional photogrammetric approaches to end-to-end #AI-based #3D reconstruction pipelines. We quantitatively evaluate the performance of #VGGT and #Fast3R using #geomatic terrestrial, #UAV & #aerial datasets isprs-archives.copernicus.org/articles/XLVII… @FBK_research

We investigates the performance of an #Android #smartphone equipped with a dual-frequency (L1, L5) #GNSS receiver and high-resolution camera as a standalone device for terrestrial #photogrammetry. isprs-archives.copernicus.org/articles/XLVII… @FBK_research #mapping

we benchmark various LSD networks - F-Clip, ULSD, L-CNN - and a #segmentation method - DeepLabV3+ for the extraction of #building #footprints from high-resolution #aerial #imagery. isprs-archives.copernicus.org/articles/XLVII… #mapping @FBK_research #AI

we showcase the capability of dual-frequency low-cost #GNSS receivers integrated with a minicomputer to measure millimeter-scale movements over extended monitoring periods paper: isprs-archives.copernicus.org/articles/XLVII… data: doi.org/10.5281/zenodo… #monitoring @FBK_research