VI3 is a model‑agnostic framework that grounds pretrained 3D foundation models (3DFMs) by using inertial measurement unit (IMU) data to provide metric scale. It initializes and preintegrates IMU readings to create a metric motion reference, which is then used to recover the scale of 3DFM outputs. The approach includes adaptable anchoring strategies for different 3DFM architectures and demonstrates scale recovery on synthetic and real aerial datasets without ground‑truth supervision.
arXiv:2609.13903v1 Announce Type: new Abstract: We study how to update a pre-built aerial scene with a newly captured, unposed ground-view sequence. The aerial scene already contains a reliable metri...
Pumpire introduces a unified benchmark for assessing metric point‑pair distance estimation in both image‑ and video‑level 3D foundation models, with or without depth priors. The benchmark uses a large, diverse dataset (pumpire‑6k) of 100 real‑world scenes, each annotated with physically measured point‑pair distances across 64 frames, totaling 6,400 frames. An evaluation protocol is provided that directly measures point‑to‑point distance errors, enabling cross‑setting comparison across 29 baseline 3D models.
Monocular depth estimation is a fundamental prerequisite for 3D reconstruction and autonomous navigation in Unmanned Aerial Vehicles (UAVs). In practical deployments, UAVs operate under highly dynamic camera poses characterized by continuous variations in height, pitch, roll, and field of view (FOV).
We present FoundationGeo, a two-stage framework that explicitly bridges relative and metric prediction via spatial calibration and principled data design. Stage 1 learns a high-fidelity, affine-invariant geometry model by initializing with DINOv3 and training on a curated 10.
FounRef is a training‑free method that refines frozen monocular foundation priors into dense metric depth by aligning them with sparse metric anchors. It validates anchors against the prior’s predictions, rejects misaligned ones, and applies a structure‑preserving solver to correct depth globally and locally while preserving fine geometry. The approach works out of the box on unseen cameras and scenes, achieving up to 24% lower depth error, 92% lower surface‑normal noise, and nearly 15× faster inference than a leading depth‑completion network.
arXiv:2601. 22054v2 Announce Type: replace-cross Abstract: Scaling has powered recent advances in vision foundation models, yet extending this paradigm to metric depth estimation remains challenging due to heterogeneous sensor noise, camera-dependent biases, and metric ambiguity in noisy cross-source 3D data.
MDE-VIO integrates learned depth priors into the VINS-Mono optimization backend to improve visual‑inertial odometry in low‑texture environments. The framework enforces affine‑invariant depth consistency and pairwise ordinal constraints while filtering unstable artifacts with variance‑based gating, keeping computation within edge‑device limits. Experiments on TartanGround and M3ED datasets show the method prevents divergence and reduces Absolute Trajectory Error by up to 28.3%.
Metric feed-forward 3D reconstruction for panoramic data remains under-explored due to the lack of large-scale panoramic RGB-D training data. We present Realsee3D, a hybrid dataset of 10K indoor scenes (1K real, 9K synthetic) with 299K panoramic viewpoints and precise metric annotations, and Argus, a feed-forward network trained on it for metric panoramic 3D reconstruction.
Metric scale monocular geometry estimation has seen significant progress through large-scale data aggregation, yet current foundation models suffer from a persistent ''scale-collapse'' phenomenon: distant landmarks and vast landscapes are metrically underestimated. We hypothesize that this performance gap stems from a training data bottleneck, where existing metric-scale datasets are hardware-constrained to homogenous vehicle-captured LiDAR or short-range indoor scans, or consist of synthetic data that lacks the semantic complexity of the physical world.
The paper introduces Epipolar Distillation (EpiDistill), a method that transfers scale‑aware geometric priors from multi‑view models to monocular depth foundation models using Rectified Stereo Tokens. By preserving epipolar attention patterns, the single‑view model maintains geometric consistency without needing multi‑view inputs during inference. Experiments show significant improvements in zero‑shot metric depth estimation on challenging datasets such as ETH3D and DIODE, and the approach consistently boosts performance of state‑of‑the‑art ViT‑based models like UniDepthV2 and DepthPro.
The paper introduces PointVGGT, a zero‑shot framework for multiview RGB‑D point cloud registration that replaces the traditional pairwise‑then‑global pipeline. It employs a foundation‑then‑refinement paradigm, using visual geometry foundation models to directly recover metrically consistent global poses and then refining them with voxelized spatial hashing and IRLS‑based bundle adjustment. Experiments on indoor, object‑centric, and outdoor datasets demonstrate superior registration accuracy and computational efficiency without any training.
GaussianDS introduces a depth‑supervised framework for 3D Gaussian Splatting that jointly optimizes RGB appearance, depth, and compact semantics from scratch. By arranging multi‑view images into a pose‑aware pseudo‑video and propagating view‑consistent masks via SAM2, the method aligns semantic lifting with geometric cues, using depth supervision and edge‑aware refinement to curb semantic drift and boundary leakage. The approach achieves state‑of‑the‑art performance on LERF and 3D‑OVS benchmarks while preserving high‑fidelity reconstruction and enabling downstream tasks such as 3D object removal.