Abstract:Computing accurate depth from multiple views is a fundamental and longstanding challenge in computer vision. However, most existing approaches do not generalize well across different domains and scene types (e.g. indoor vs. outdoor). Training a general-purpose multi-view stereo model is challenging and raises several questions, e.g. how to best make use of transformer-based architectures, how to incorporate additional metadata when there is a variable number of input views, and how to estimate the range of valid depths which can vary considerably across different scenes and is typically not known a priori? To address these issues, we introduce MVSA, a novel and versatile Multi-View Stereo architecture that aims to work Anywhere by generalizing across diverse domains and depth ranges. MVSA combines monocular and multi-view cues with an adaptive cost volume to deal with scale-related issues. We demonstrate state-of-the-art zero-shot depth estimation on the Robust Multi-View Depth Benchmark, surpassing existing multi-view stereo and monocular baselines.
Abstract:We present AnyCalib, a method for calibrating the intrinsic parameters of a camera from a single in-the-wild image, that is agnostic to the camera model. Current methods are predominantly tailored to specific camera models and/or require extrinsic cues, such as the direction of gravity, to be visible in the image. In contrast, we argue that the perspective and distortion cues inherent in images are sufficient for model-agnostic camera calibration. To demonstrate this, we frame the calibration process as the regression of the rays corresponding to each pixel. We show, for the first time, that this intermediate representation allows for a closed-form recovery of the intrinsics for a wide range of camera models, including but not limited to: pinhole, Brown-Conrady and Kannala-Brandt. Our approach also applies to edited -- cropped and stretched -- images. Experimentally, we demonstrate that AnyCalib consistently outperforms alternative methods, including 3D foundation models, despite being trained on orders of magnitude less data. Code is available at https://github.com/javrtg/AnyCalib.
Abstract:Works based on localization and mapping do not exploit the inherent semantic-relational information from the environment for faster and efficient management and optimization of the robot poses and its map elements, often leading to pose and map inaccuracies and computational inefficiencies in large scale environments. 3D scene graph representations which distributes the environment in an hierarchical manner can be exploited to enhance the management/optimization of underlying robot poses and its map. In this direction, we present our work Situational Graphs 2.0, which leverages the hierarchical structure of indoor scenes for efficient data management and optimization. Our algorithm begins by constructing a situational graph that organizes the environment into four layers: Keyframes, Walls, Rooms, and Floors. Our first novelty lies in the front-end which includes a floor detection module capable of identifying stairways and assigning a floor-level semantic-relations to the underlying layers. This floor-level semantic enables a floor-based loop closure strategy, rejecting false-positive loop closures in visually similar areas on different floors. Our second novelty is in exploiting the hierarchy for an improved optimization. It consists of: (1) local optimization, optimizing a window of recent keyframes and their connected components, (2) floor-global optimization, which focuses only on keyframes and their connections within the current floor during loop closures, and (3) room-local optimization, marginalizing redundant keyframes that share observations within the room. We validate our algorithm extensively in different real multi-floor environments. Our approach can demonstrate state-of-art-art results in large scale multi-floor environments creating hierarchical maps while bounding the computational complexity where several baseline works fail to execute efficiently.
Abstract:Metric depth estimation from visual sensors is crucial for robots to perceive, navigate, and interact with their environment. Traditional range imaging setups, such as stereo or structured light cameras, face hassles including calibration, occlusions, and hardware demands, with accuracy limited by the baseline between cameras. Single- and multi-view monocular depth offers a more compact alternative, but is constrained by the unobservability of the metric scale. Light field imaging provides a promising solution for estimating metric depth by using a unique lens configuration through a single device. However, its application to single-view dense metric depth is under-addressed mainly due to the technology's high cost, the lack of public benchmarks, and proprietary geometrical models and software. Our work explores the potential of focused plenoptic cameras for dense metric depth. We propose a novel pipeline that predicts metric depth from a single plenoptic camera shot by first generating a sparse metric point cloud using machine learning, which is then used to scale and align a dense relative depth map regressed by a foundation depth model, resulting in dense metric depth. To validate it, we curated the Light Field & Stereo Image Dataset (LFS) of real-world light field images with stereo depth labels, filling a current gap in existing resources. Experimental results show that our pipeline produces accurate metric depth predictions, laying a solid groundwork for future research in this field.
Abstract:Evaluation is critical to both developing and tuning Structure from Motion (SfM) and Visual SLAM (VSLAM) systems, but is universally reliant on high-quality geometric ground truth -- a resource that is not only costly and time-intensive but, in many cases, entirely unobtainable. This dependency on ground truth restricts SfM and SLAM applications across diverse environments and limits scalability to real-world scenarios. In this work, we propose a novel ground-truth-free (GTF) evaluation methodology that eliminates the need for geometric ground truth, instead using sensitivity estimation via sampling from both original and noisy versions of input images. Our approach shows strong correlation with traditional ground-truth-based benchmarks and supports GTF hyperparameter tuning. Removing the need for ground truth opens up new opportunities to leverage a much larger number of dataset sources, and for self-supervised and online tuning, with the potential for a data-driven breakthrough analogous to what has occurred in generative AI.
Abstract:This paper presents the first Open-Vocabulary Online 3D semantic SLAM pipeline, that we denote as OVO-SLAM. Our primary contribution is in the pipeline itself, particularly in the mapping thread. Given a set of posed RGB-D frames, we detect and track 3D segments, which we describe using CLIP vectors, calculated through a novel aggregation from the viewpoints where these 3D segments are observed. Notably, our OVO-SLAM pipeline is not only faster but also achieves better segmentation metrics compared to offline approaches in the literature. Along with superior segmentation performance, we show experimental results of our contributions integrated with Gaussian-SLAM, being the first ones demonstrating end-to-end open-vocabulary online 3D reconstructions without relying on ground-truth camera poses or scene geometry.
Abstract:While visual SLAM systems are well studied and achieve impressive results in indoor and urban settings, natural, outdoor and open-field environments are much less explored and still present relevant research challenges. Visual navigation and local mapping have shown a relatively good performance in open-field environments. However, globally consistent mapping and long-term localization still depend on the robustness of loop detection and closure, for which the literature is scarce. In this work we propose a novel method to pave the way towards robust loop detection in open fields, particularly in agricultural settings, based on local feature search and stereo geometric refinement, with a final stage of relative pose estimation. Our method consistently achieves good loop detections, with a median error of 15cm. We aim to characterize open fields as a novel environment for loop detection, understanding the limitations and problems that arise when dealing with them.
Abstract:Having prior knowledge of an environment boosts the localization and mapping accuracy of robots. Several approaches in the literature have utilized architectural plans in this regard. However, almost all of them overlook the deviations between actual as-built environments and as-planned architectural designs, introducing bias in the estimations. To address this issue, we present a novel localization and mapping method denoted as deviations-informed Situational Graphs or diS-Graphs that integrates prior knowledge from architectural plans even in the presence of deviations. It is based on Situational Graphs (S-Graphs) that merge geometric models of the environment with 3D scene graphs into a multi-layered jointly optimizable factor graph. Our diS-Graph extracts information from architectural plans by first modeling them as a hierarchical factor graph, which we will call an Architectural Graph (A-Graph). While the robot explores the real environment, it estimates an S-Graph from its onboard sensors. We then use a novel matching algorithm to register the A-Graph and S-Graph in the same reference, and merge both of them with an explicit model of deviations. Finally, an alternating graph optimization strategy allows simultaneous global localization and mapping, as well as deviation estimation between both the A-Graph and the S-Graph. We perform several experiments in simulated and real datasets in the presence of deviations. On average, our diS-Graphs outperforms the baselines by a margin of approximately 43% in simulated environments and by 7% in real environments, while being able to estimate deviations up to 35 cm and 15 degrees.
Abstract:We propose three novel metrics for evaluating the accuracy of a set of estimated camera poses given the ground truth: Translation Alignment Score (TAS), Rotation Alignment Score (RAS), and Pose Alignment Score (PAS). The TAS evaluates the translation accuracy independently of the rotations, and the RAS evaluates the rotation accuracy independently of the translations. The PAS is the average of the two scores, evaluating the combined accuracy of both translations and rotations. The TAS is computed in four steps: (1) Find the upper quartile of the closest-pair-distances, $d$. (2) Align the estimated trajectory to the ground truth using a robust registration method. (3) Collect all distance errors and obtain the cumulative frequencies for multiple thresholds ranging from $0.01d$ to $d$ with a resolution $0.01d$. (4) Add up these cumulative frequencies and normalize them such that the theoretical maximum is 1. The TAS has practical advantages over the existing metrics in that (1) it is robust to outliers and collinear motion, and (2) there is no need to adjust parameters on different datasets. The RAS is computed in a similar manner to the TAS and is also shown to be more robust against outliers than the existing rotation metrics. We verify our claims through extensive simulations and provide in-depth discussion of the strengths and weaknesses of the proposed metrics.
Abstract:Visual Place Recognition (VPR) plays a critical role in many localization and mapping pipelines. It consists of retrieving the closest sample to a query image, in a certain embedding space, from a database of geotagged references. The image embedding is learned to effectively describe a place despite variations in visual appearance, viewpoint, and geometric changes. In this work, we formulate how limitations in the Geographic Distance Sensitivity of current VPR embeddings result in a high probability of incorrectly sorting the top-k retrievals, negatively impacting the recall. In order to address this issue in single-stage VPR, we propose a novel mining strategy, CliqueMining, that selects positive and negative examples by sampling cliques from a graph of visually similar images. Our approach boosts the sensitivity of VPR embeddings at small distance ranges, significantly improving the state of the art on relevant benchmarks. In particular, we raise recall@1 from 75% to 82% in MSLS Challenge, and from 76% to 90% in Nordland. Models and code are available at https://github.com/serizba/cliquemining.