M3GA-Wild is a new benchmark for multi-modal, multi-session ground-to-aerial place recognition in forests, combining synchronized RGB imagery and LiDAR from ground traversals with high‑resolution aerial imagery and multi‑altitude LiDAR over 370 hectares. The dataset includes accurate geo‑referenced 6‑DoF poses and spans 36 km of forest traversals, enabling systematic evaluation of visual, LiDAR, cross‑modal, and multi‑modal methods. Baseline experiments show LiDAR outperforms vision‑only approaches under severe viewpoint changes, while current multi‑modal fusion offers limited gains due to poor cross‑modal alignment, highlighting challenges in cross‑platform localisation and domain gaps.
We present M3GA-Wild, the first benchmark for multi-modal, multi-session ground-to-aerial place recognition in forests. M3GA-Wild unifies and extends existing forest localisation datasets, providing a...
Cross-modal place recognition (CMPR) aims to identify the same location across heterogeneous sensing modalities, such as vision and LiDAR. Existing methods commonly bridge the modality gap using complex alignment modules, multi-stage training, or full fine-tuning of pretrained backbones.
DXPR is a depth‑based cross‑modal place recognition framework that matches monocular camera queries to a LiDAR map using a single vision foundation model backbone. By converting both modalities into a unified depth image representation, DXPR learns modality‑invariant global descriptors without modality‑specific encoders. A geometry‑aware overlap miner refines pairwise metric learning by computing pixel‑level overlap scores, and extensive tests on KITTI and Boreas show strong performance across seasons, weather, and day/night conditions, outperforming prior CMPR baselines.
arXiv:2606. 30576v1 Announce Type: cross Abstract: Cross-view object geo-localization (CVOGL) aims to locate a target object from a query view (e.
OpenCVL is a large, open dataset for fine-grained cross-view localization, comprising 617,388 ground‑aerial image pairs from 41 European cities. It blends high‑end sensor data with diverse in‑the‑wild images and includes a curation framework to correct pose annotations, enabling reliable evaluation. The dataset also offers cross‑area and snowy test sets to probe generalization, and experiments show that adding noisy in‑the‑wild data improves model performance on clean tests.
arXiv:2602.14929v2 Announce Type: replace Abstract: Aligning ground-level imagery with geo-registered satellite maps is crucial for mapping, navigation, and situational awareness, yet remains challen...
M3GD introduces a multimodal representation that fuses pre‑trained 2D image and 3D LiDAR foundation models for robotic novel view synthesis, avoiding the need for a separate cross‑modal translator. By projecting LiDAR onto the image latent grid and injecting the resulting geometry‑aware packets via a lightweight residual adapter, the method enhances both RGB and depth synthesis on the GrandTour dataset compared to an image‑only baseline. Ablation studies confirm that pixel‑aligned LiDAR content drives the performance gains, and real‑world deployment on a ground robot demonstrates a tunable quality–cost trade‑off.
Global visual localization of unmanned aerial vehicles (UAVs) using remote-sensing reference maps has attracted increasing attention. However, acquisition-time and imaging-platform differences between UAV and reference imagery induce substantial cross-domain appearance and viewpoint shifts, challenging robust six-degree-of-freedom (6-DoF) pose estimation.
arXiv:2608. 09270v1 Announce Type: cross Abstract: Fine-grained cross-modal understanding in drone views is essential for aerial vision-language navigation.
UniGeo is a multimodal large language model designed for text-guided drone geo‑localization, enabling the identification of target regions in large image galleries from natural‑language descriptions. It integrates geo‑semantic understanding, cross‑view semantic generation, and candidate‑level verification within a shared vision‑language framework, establishing stable correspondences among local scene elements, spatial relations, and language. A multi‑stage training strategy progressively refines geo‑semantic learning, cross‑view mapping, and fine‑grained verification, yielding significant performance gains on GeoText‑1652, with R@10 and mAP improvements of 13.59 and 2.83 percentage points respectively.
arXiv:2607.20116v2 Announce Type: replace Abstract: Global visual localization of unmanned aerial vehicles (UAVs) using remote-sensing reference maps has attracted increasing attention. However, diff...
arXiv:2607. 08711v1 Announce Type: cross Abstract: Accurate 3D terrain maps are essential for emergency response when assessing wildfire hazards.