no code implementations • 28 Mar 2024 • Ganlin Zhang, Erik Sandström, Youmin Zhang, Manthan Patel, Luc van Gool, Martin R. Oswald
To alleviate this issue, with the aid of a monocular depth estimator, we introduce a novel DSPO layer for bundle adjustment which optimizes the pose and depth of keyframes along with the scale of the monocular depth.
1 code implementation • 20 Feb 2024 • Fabio Tosi, Youmin Zhang, Ziren Gong, Erik Sandström, Stefano Mattoccia, Martin R. Oswald, Matteo Poggi
Over the past two decades, research in the field of Simultaneous Localization and Mapping (SLAM) has undergone a significant evolution, highlighting its critical role in enabling autonomous exploration of unknown environments.
no code implementations • 14 Feb 2024 • Lorenzo Liso, Erik Sandström, Vladimir Yugay, Luc van Gool, Martin R. Oswald
Neural RGBD SLAM techniques have shown promise in dense Simultaneous Localization And Mapping (SLAM), yet face challenges such as error accumulation during camera tracking resulting in distorted maps.
no code implementations • 19 Jan 2024 • Zuoyue Li, Zhenqiang Li, Zhaopeng Cui, Marc Pollefeys, Martin R. Oswald
Directly generating scenes from satellite imagery offers exciting possibilities for integration into applications like games and map services.
no code implementations • 8 Jan 2024 • Casimir Feldmann, Niall Siegenheim, Nikolas Hars, Lovro Rabuzin, Mert Ertugrul, Luca Wolfart, Marc Pollefeys, Zuria Bauer, Martin R. Oswald
In the case of MDE models for autonomous driving, this issue is exacerbated by the linearity of the captured data trajectories.
no code implementations • 15 Dec 2023 • Weijie Wei, Fatemeh Karimi Nejadasl, Theo Gevers, Martin R. Oswald
The scarcity of annotated data in LiDAR point cloud understanding hinders effective representation learning.
1 code implementation • 7 Dec 2023 • Osman Ülger, Maksymilian Kulicki, Yuki Asano, Martin R. Oswald
Open-Vocabulary Segmentation (OVS) methods are capable of performing semantic segmentation without relying on a fixed vocabulary, and in some cases, they operate without the need for training or fine-tuning.
no code implementations • 6 Dec 2023 • Vladimir Yugay, Yue Li, Theo Gevers, Martin R. Oswald
We present a dense simultaneous localization and mapping (SLAM) method that uses 3D Gaussians as a scene representation.
no code implementations • 30 Nov 2023 • Aritra Bhowmik, Martin R. Oswald, Pascal Mettes, Cees G. M. Snoek
For proposal regression, we solve a simpler problem where we regress to the area of intersection between proposal and ground truth.
no code implementations • 29 Nov 2023 • Silvan Weder, Francis Engelmann, Johannes L. Schönberger, Akihito Seki, Marc Pollefeys, Martin R. Oswald
Using these main contributions, our method can enable scenarios with real-time constraints and can scale to arbitrary scene sizes by processing and updating the scene only in a local region defined by the new measurement.
1 code implementation • 11 Oct 2023 • Osman Ülger, Yu Wang, Ysbrand Galama, Sezer Karaoglu, Theo Gevers, Martin R. Oswald
Humans have a remarkable ability to perceive and reason about the world around them by understanding the relationships between objects.
no code implementations • 9 Oct 2023 • Duy-Kien Nguyen, Martin R. Oswald, Cees G. M. Snoek
The ability to detect objects in images at varying scales has played a pivotal role in the design of modern object detectors.
no code implementations • 30 Sep 2023 • Jose Andres Millan-Romera, Hriday Bavle, Muhammad Shaheer, Martin R. Oswald, Holger Voos, Jose Luis Sanchez-Lopez
Concretely, our previous work, Situational Graphs (S-Graphs+), a pioneer in jointly leveraging semantic relationships in the factor optimization process, relies on semantic entities such as Planes and Rooms, whose relationship is mathematically defined.
1 code implementation • 29 Sep 2023 • Weijie Wei, Martin R. Oswald, Fatemeh Karimi Nejadasl, Theo Gevers
To leverage the different properties of each branch, we employ a geometry-aware fusion module that is learned to combine the results of each branch.
no code implementations • 5 Aug 2023 • Florentin Liebmann, Marco von Atzigen, Dominik Stütz, Julian Wolf, Lukas Zingg, Daniel Suter, Laura Leoty, Hooman Esfandiari, Jess G. Snedeker, Martin R. Oswald, Marc Pollefeys, Mazda Farshad, Philipp Fürnstahl
An intuitive surgical guidance is provided thanks to the integration into an augmented reality based navigation system.
1 code implementation • 29 Jun 2023 • David Recasens, Martin R. Oswald, Marc Pollefeys, Javier Civera
Estimating camera motion in deformable scenes poses a complex and open research challenge.
1 code implementation • 19 Jun 2023 • Erik Sandström, Kevin Ta, Luc van Gool, Martin R. Oswald
We present an uncertainty learning framework for dense neural simultaneous localization and mapping (SLAM).
1 code implementation • 8 Jun 2023 • Duy-Kien Nguyen, Vaibhav Aggarwal, Yanghao Li, Martin R. Oswald, Alexander Kirillov, Cees G. M. Snoek, Xinlei Chen
In this work, we explore regions as a potential visual analogue of words for self-supervised image representation learning.
no code implementations • 3 May 2023 • Cathrin Elich, Iro Armeni, Martin R. Oswald, Marc Pollefeys, Joerg Stueckler
Our approach compares favorably to previous state-of-the-art object-level matching approaches and achieves improved performance over a pure keypoint-based approach for large view-point changes.
no code implementations • ICCV 2023 • Denys Rozumnyi, Jiri Matas, Marc Pollefeys, Vittorio Ferrari, Martin R. Oswald
We argue that this representation is limited and instead propose to guide and improve 2D tracking with an explicit object representation, namely the textured 3D shape and 6DoF pose in each video frame.
2 code implementations • ICCV 2023 • Erik Sandström, Yue Li, Luc van Gool, Martin R. Oswald
We propose a dense neural simultaneous localization and mapping (SLAM) approach for monocular RGBD input which anchors the features of a neural scene representation in a point cloud that is iteratively generated in an input-dependent data-driven manner.
no code implementations • ICCV 2023 • Yiming Zhao, Denys Rozumnyi, Jie Song, Otmar Hilliges, Marc Pollefeys, Martin R. Oswald
The key idea is to tackle the inverse problem of image deblurring by modeling the forward problem with a 3D human model, a texture map, and a sequence of poses to describe human motion.
no code implementations • 7 Feb 2023 • Zihan Zhu, Songyou Peng, Viktor Larsson, Zhaopeng Cui, Martin R. Oswald, Andreas Geiger, Marc Pollefeys
Neural implicit representations have recently become popular in simultaneous localization and mapping (SLAM), especially in dense visual SLAM.
no code implementations • ICCV 2023 • Aritra Bhowmik, Yu Wang, Nora Baka, Martin R. Oswald, Cees G. M. Snoek
Contrary to existing methods, which learn objects and relations separately, our key idea is to learn the object-relation distribution jointly.
1 code implementation • CVPR 2023 • Rémi Pautrat, Daniel Barath, Viktor Larsson, Martin R. Oswald, Marc Pollefeys
Their learned counterparts are more repeatable and can handle challenging images, but at the cost of a lower accuracy and a bias towards wireframe lines.
no code implementations • 5 Oct 2022 • Mathias Vetsch, Sandro Lombardi, Marc Pollefeys, Martin R. Oswald
The generation of triangle meshes from point clouds, i. e. meshing, is a core task in computer graphics and computer vision.
no code implementations • 23 Jul 2022 • Zuoyue Li, Tianxing Fan, Zhenqiang Li, Zhaopeng Cui, Yoichi Sato, Marc Pollefeys, Martin R. Oswald
We introduce a scalable framework for novel view synthesis from RGB-D images with largely incomplete scene coverage.
1 code implementation • 7 Apr 2022 • Erik Sandström, Martin R. Oswald, Suryansh Kumar, Silvan Weder, Fisher Yu, Cristian Sminchisescu, Luc van Gool
Multi-sensor depth fusion is able to substantially improve the robustness and accuracy of 3D reconstruction methods, but existing techniques are not robust enough to handle sensors which operate with diverse value ranges as well as noise and outlier statistics.
no code implementations • 29 Mar 2022 • Christian Sigg, Flavia Cavallaro, Tobias Günther, Martin R. Oswald
This is challenging, because photographic visualizations of weather forecasts should look real, be free of obvious artifacts, and should match the predicted weather conditions.
1 code implementation • CVPR 2022 • Zihan Zhu, Songyou Peng, Viktor Larsson, Weiwei Xu, Hujun Bao, Zhaopeng Cui, Martin R. Oswald, Marc Pollefeys
Neural implicit representations have recently shown encouraging results in various domains, including promising progress in simultaneous localization and mapping (SLAM).
no code implementations • 22 Dec 2021 • Zuria Bauer, Zuoyue Li, Sergio Orts-Escolano, Miguel Cazorla, Marc Pollefeys, Martin R. Oswald
Building upon the recent progress in novel view synthesis, we propose its application to improve monocular depth estimation.
Ranked #27 on Monocular Depth Estimation on KITTI Eigen split
1 code implementation • CVPR 2022 • Denys Rozumnyi, Martin R. Oswald, Vittorio Ferrari, Marc Pollefeys
We propose a method for jointly estimating the 3D motion, 3D shape, and appearance of highly motion-blurred objects from a video.
1 code implementation • CVPR 2022 • Duy-Kien Nguyen, Jihong Ju, Olaf Booij, Martin R. Oswald, Cees G. M. Snoek
Specifically, we present BoxeR, short for Box Transformer, which attends to a set of boxes by predicting their transformation from a reference window on an input feature map.
no code implementations • 13 Oct 2021 • Qingshan Xu, Martin R. Oswald, Wenbing Tao, Marc Pollefeys, Zhaopeng Cui
However, existing recurrent methods only model the local dependencies in the depth domain, which greatly limits the capability of capturing the global scene context along the depth dimension.
no code implementations • 31 Aug 2021 • Michael Seeber, Roi Poranne, Marc Polleyfeys, Martin R. Oswald
Estimating 3D hand meshes from RGB images robustly is a highly desirable task, made challenging due to the numerous degrees of freedom, and issues such as self similarity and occlusions.
1 code implementation • 11 Aug 2021 • Davide Menini, Suryansh Kumar, Martin R. Oswald, Erik Sandstrom, Cristian Sminchisescu, Luc van Gool
This paper presents a real-time online vision framework to jointly recover an indoor scene's 3D structure and semantic label.
1 code implementation • NeurIPS 2021 • Denys Rozumnyi, Martin R. Oswald, Vittorio Ferrari, Marc Pollefeys
We address the novel task of jointly reconstructing the 3D shape, texture, and motion of an object from a single motion-blurred image.
1 code implementation • CVPR 2021 • Rémi Pautrat, Juan-Ting Lin, Viktor Larsson, Martin R. Oswald, Marc Pollefeys
We thus hereby introduce the first joint detection and description of line segments in a single deep network.
no code implementations • 31 Dec 2020 • Ayça Takmaz, Danda Pani Paudel, Thomas Probst, Ajad Chhatkuli, Martin R. Oswald, Luc van Gool
In this work, we present an unsupervised monocular framework for dense depth estimation of dynamic scenes, which jointly reconstructs rigid and non-rigid parts without explicitly modelling the camera motion.
1 code implementation • CVPR 2021 • Marko Mihajlovic, Silvan Weder, Marc Pollefeys, Martin R. Oswald
We present DeepSurfels, a novel hybrid scene representation for geometry and appearance information.
1 code implementation • ICCV 2021 • Denys Rozumnyi, Jiri Matas, Filip Sroubek, Marc Pollefeys, Martin R. Oswald
Compared to other methods, such as deblatting, the inference is of several orders of magnitude faster and allows applications such as real-time fast moving object detection and retrieval in large video collections.
no code implementations • ICCV 2021 • Zuoyue Li, Zhenqiang Li, Zhaopeng Cui, Rongjun Qin, Marc Pollefeys, Martin R. Oswald
For geometrical and temporal consistency, our approach explicitly creates a 3D point cloud representation of the scene and maintains dense 3D-2D correspondences across frames that reflect the geometric scene configuration inferred from the satellite view.
5 code implementations • CVPR 2021 • Denys Rozumnyi, Martin R. Oswald, Vittorio Ferrari, Jiri Matas, Marc Pollefeys
We propose a method that, given a single image with its estimated background, outputs the object's appearance and position in a series of sub-frames as if captured by a high-speed camera (i. e. temporal super-resolution).
Ranked #1 on Video Super-Resolution on Falling Objects
1 code implementation • CVPR 2021 • Silvan Weder, Johannes L. Schönberger, Marc Pollefeys, Martin R. Oswald
We present a novel online depth map fusion approach that learns depth map aggregation in a latent feature space.
no code implementations • 8 Oct 2020 • Cathrin Elich, Martin R. Oswald, Marc Pollefeys, Joerg Stueckler
Our approach learns to decompose images of synthetic scenes with multiple objects on a planar surface into its constituent scene objects and to infer their 3D properties from a single view.
no code implementations • 28 Sep 2020 • Cathrin Elich, Martin R. Oswald, Marc Pollefeys, Joerg Stueckler
By differentiable rendering, we train our model to decompose scenes self-supervised from RGB-D images.
no code implementations • 22 Sep 2020 • Ivan Tishchenko, Sandro Lombardi, Martin R. Oswald, Marc Pollefeys
Most of the current scene flow methods choose to model scene flow as a per point translation vector without differentiating between static and dynamic components of 3D motion.
no code implementations • 31 Jul 2020 • Audrey Richard, Ian Cherabier, Martin R. Oswald, Marc Pollefeys, Konrad Schindler
We present a novel 3D shape completion method that operates directly on unstructured point clouds, thus avoiding resource-intensive data structures like voxel grids.
1 code implementation • ECCV 2020 • Rémi Pautrat, Viktor Larsson, Martin R. Oswald, Marc Pollefeys
To be invariant, or not to be invariant: that is the question formulated in this work about local descriptors.
2 code implementations • 17 Jan 2020 • Lucas Teixeira, Martin R. Oswald, Marc Pollefeys, Margarita Chli
In this paper, we propose a depth completion and uncertainty estimation approach that better handles the challenges of aerial platforms, such as large viewpoint and depth variations, and limited computing resources.
no code implementations • 14 Jan 2020 • Audrey Richard, Ian Cherabier, Martin R. Oswald, Vagia Tsiminaki, Marc Pollefeys, Konrad Schindler
We present a super-resolution method capable of creating a high-resolution texture map for a virtual 3D object from a set of lower-resolution images of that object.
2 code implementations • CVPR 2020 • Silvan Weder, Johannes L. Schönberger, Marc Pollefeys, Martin R. Oswald
To this end, we present a novel real-time capable machine learning-based method for depth map fusion.
no code implementations • 2 Sep 2019 • Denys Rozumnyi, Ian Cherabier, Marc Pollefeys, Martin R. Oswald
Our method learns sensor or algorithm properties jointly with semantic depth fusion and scene completion and can also be used as an expert system, e. g. to unify the strengths of various photometric stereo algorithms.
no code implementations • ICCV 2019 • Jean Lahoud, Bernard Ghanem, Marc Pollefeys, Martin R. Oswald
The second goal is to learn instance information by densely estimating directional information of the instance's center of mass for each voxel.
Ranked #2 on 3D Semantic Instance Segmentation on ScanNetV2
no code implementations • ECCV 2018 • Ian Cherabier, Johannes L. Schonberger, Martin R. Oswald, Marc Pollefeys, Andreas Geiger
In contrast to existing variational methods for semantic 3D reconstruction, our model is end-to-end trainable and captures more complex dependencies between the semantic labels and the 3D geometry.
no code implementations • CVPR 2018 • Pablo Speciale, Danda P. Paudel, Martin R. Oswald, Hayko Riemenschneider, Luc van Gool, Marc Pollefeys
We propose a novel method for the geometric registration of semantically labeled regions.
no code implementations • CVPR 2017 • Fabio Maninchedda, Martin R. Oswald, Marc Pollefeys
We present a method for the fast 3D face reconstruction of people wearing glasses.
no code implementations • CVPR 2017 • Pablo Speciale, Danda Pani Paudel, Martin R. Oswald, Till Kroeger, Luc van Gool, Marc Pollefeys
While randomized methods like RANSAC are fast, they do not guarantee global optimality and fail to manage large amounts of outliers.
no code implementations • ICCV 2017 • Maros Blaha, Mathias Rothermel, Martin R. Oswald, Torsten Sattler, Audrey Richard, Jan D. Wegner, Marc Pollefeys, Konrad Schindler
We present a method to jointly refine the geometry and semantic segmentation of 3D surface meshes.
no code implementations • ICCV 2015 • Mohamed Souiai, Martin R. Oswald, Youngwook Kee, Junmo Kim, Marc Pollefeys, Daniel Cremers
Despite their enormous success in solving hard combinatorial problems, convex relaxation approaches often suffer from the fact that the computed solutions are far from binary and that subsequent heuristic binarization may substantially degrade the quality of computed solutions.