Monocular depth reconstruction of complex and dynamic scenes is a highly challenging problem. While for rigid scenes learning-based methods have been offering promising results even in unsupervised cases, there exists little to no literature addressing the same for dynamic and deformable scenes. 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. Using dense correspondences, we derive a training objective that aims to opportunistically preserve pairwise distances between reconstructed 3D points. In this process, the dense depth map is learned implicitly using the as-rigid-as-possible hypothesis. Our method provides promising results, demonstrating its capability of reconstructing 3D from challenging videos of non-rigid scenes. Furthermore, the proposed method also provides unsupervised motion segmentation results as an auxiliary output.
相關內容
Despite the recent developments in 3D Face Reconstruction from occluded and noisy face images, the performance is still unsatisfactory. One of the main challenges is to handle moderate to heavy occlusions in the face images. In addition, the noise in the face images inhibits the correct capture of facial attributes, thus needing to be reliably addressed. Moreover, most existing methods rely on additional dependencies, posing numerous constraints over the training procedure. Therefore, we propose a Self-Supervised RObustifying GUidancE (ROGUE) framework to obtain robustness against occlusions and noise in the face images. The proposed network contains 1) the Guidance Pipeline to obtain the 3D face coefficients for the clean faces, and 2) the Robustification Pipeline to acquire the consistency between the estimated coefficients for occluded or noisy images and the clean counterpart. The proposed image- and feature-level loss functions aid the ROGUE learning process without posing additional dependencies. On the three variations of the test dataset of CelebA: rational occlusions, delusional occlusions, and noisy face images, our method outperforms the current state-of-the-art method by large margins (e.g., for the shape-based 3D vertex errors, a reduction from 0.146 to 0.048 for rational occlusions, from 0.292 to 0.061 for delusional occlusions and from 0.269 to 0.053 for the noise in the face images), demonstrating the effectiveness of the proposed approach.
This paper presents a real-time online vision framework to jointly recover an indoor scene's 3D structure and semantic label. Given noisy depth maps, a camera trajectory, and 2D semantic labels at train time, the proposed deep neural network based approach learns to fuse the depth over frames with suitable semantic labels in the scene space. Our approach exploits the joint volumetric representation of the depth and semantics in the scene feature space to solve this task. For a compelling online fusion of the semantic labels and geometry in real-time, we introduce an efficient vortex pooling block while dropping the use of routing network in online depth fusion to preserve high-frequency surface details. We show that the context information provided by the semantics of the scene helps the depth fusion network learn noise-resistant features. Not only that, it helps overcome the shortcomings of the current online depth fusion method in dealing with thin object structures, thickening artifacts, and false surfaces. Experimental evaluation on the Replica dataset shows that our approach can perform depth fusion at 37 and 10 frames per second with an average reconstruction F-score of 88% and 91%, respectively, depending on the depth map resolution. Moreover, our model shows an average IoU score of 0.515 on the ScanNet 3D semantic benchmark leaderboard.
Novel view synthesis for humans in motion is a challenging computer vision problem that enables applications such as free-viewpoint video. Existing methods typically use complex setups with multiple input views, 3D supervision, or pre-trained models that do not generalize well to new identities. Aiming to address these limitations, we present a novel view synthesis framework to generate realistic renders from unseen views of any human captured from a single-view sensor with sparse RGB-D, similar to a low-cost depth camera, and without actor-specific models. We propose an architecture to learn dense features in novel views obtained by sphere-based neural rendering, and create complete renders using a global context inpainting model. Additionally, an enhancer network leverages the overall fidelity, even in occluded areas from the original view, producing crisp renders with fine details. We show our method generates high-quality novel views of synthetic and real human actors given a single sparse RGB-D input. It generalizes to unseen identities, new poses and faithfully reconstructs facial expressions. Our approach outperforms prior human view synthesis methods and is robust to different levels of input sparsity.
A video autoencoder is proposed for learning disentan- gled representations of 3D structure and camera pose from videos in a self-supervised manner. Relying on temporal continuity in videos, our work assumes that the 3D scene structure in nearby video frames remains static. Given a sequence of video frames as input, the video autoencoder extracts a disentangled representation of the scene includ- ing: (i) a temporally-consistent deep voxel feature to represent the 3D structure and (ii) a 3D trajectory of camera pose for each frame. These two representations will then be re-entangled for rendering the input video frames. This video autoencoder can be trained directly using a pixel reconstruction loss, without any ground truth 3D or camera pose annotations. The disentangled representation can be applied to a range of tasks, including novel view synthesis, camera pose estimation, and video generation by motion following. We evaluate our method on several large- scale natural video datasets, and show generalization results on out-of-domain images.
In the last decade, numerous supervised deep learning approaches requiring large amounts of labeled data have been proposed for visual-inertial odometry (VIO) and depth map estimation. To overcome the data limitation, self-supervised learning has emerged as a promising alternative, exploiting constraints such as geometric and photometric consistency in the scene. In this study, we introduce a novel self-supervised deep learning-based VIO and depth map recovery approach (SelfVIO) using adversarial training and self-adaptive visual-inertial sensor fusion. SelfVIO learns to jointly estimate 6 degrees-of-freedom (6-DoF) ego-motion and a depth map of the scene from unlabeled monocular RGB image sequences and inertial measurement unit (IMU) readings. The proposed approach is able to perform VIO without the need for IMU intrinsic parameters and/or the extrinsic calibration between the IMU and the camera. estimation and single-view depth recovery network. We provide comprehensive quantitative and qualitative evaluations of the proposed framework comparing its performance with state-of-the-art VIO, VO, and visual simultaneous localization and mapping (VSLAM) approaches on the KITTI, EuRoC and Cityscapes datasets. Detailed comparisons prove that SelfVIO outperforms state-of-the-art VIO approaches in terms of pose estimation and depth recovery, making it a promising approach among existing methods in the literature.
Vision-based Simultaneous Localization And Mapping (VSLAM) is a mature problem in Robotics. Most VSLAM systems are feature based methods, which are robust and present high accuracy, but yield sparse maps with limited application for further navigation tasks. Most recently, direct methods which operate directly on image intensity have been introduced, capable of reconstructing richer maps at the cost of higher processing power. In this work, an edge-based monocular SLAM system (SE-SLAM) is proposed as a middle point: edges present good localization as point features, while enabling a structural semidense map reconstruction. However, edges are not easy to associate, track and optimize over time, as they lack descriptors and biunivocal correspondence, unlike point features. To tackle these issues, this paper presents a method to match edges between frames in a consistent manner; a feasible strategy to solve the optimization problem, since its size rapidly increases when working with edges; and the use of non-linear optimization techniques. The resulting system achieves comparable precision to state of the art feature-based and dense/semi-dense systems, while inherently building a structural semi-dense reconstruction of the environment, providing relevant structure data for further navigation algorithms. To achieve such accuracy, state of the art non-linear optimization is needed, over a continuous feed of 10000 edgepoints per frame, to optimize the full semi-dense output. Despite its heavy processing requirements, the system achieves near to real-time operation, thanks to a custom built solver and parallelization of its key stages. In order to encourage further development of edge-based SLAM systems, SE-SLAM source code will be released as open source.
In this paper, we proposed a new deep learning based dense monocular SLAM method. Compared to existing methods, the proposed framework constructs a dense 3D model via a sparse to dense mapping using learned surface normals. With single view learned depth estimation as prior for monocular visual odometry, we obtain both accurate positioning and high quality depth reconstruction. The depth and normal are predicted by a single network trained in a tightly coupled manner.Experimental results show that our method significantly improves the performance of visual tracking and depth prediction in comparison to the state-of-the-art in deep monocular dense SLAM.
Single-image piece-wise planar 3D reconstruction aims to simultaneously segment plane instances and recover 3D plane parameters from an image. Most recent approaches leverage convolutional neural networks (CNNs) and achieve promising results. However, these methods are limited to detecting a fixed number of planes with certain learned order. To tackle this problem, we propose a novel two-stage method based on associative embedding, inspired by its recent success in instance segmentation. In the first stage, we train a CNN to map each pixel to an embedding space where pixels from the same plane instance have similar embeddings. Then, the plane instances are obtained by grouping the embedding vectors in planar regions via an efficient mean shift clustering algorithm. In the second stage, we estimate the parameter for each plane instance by considering both pixel-level and instance-level consistencies. With the proposed method, we are able to detect an arbitrary number of planes. Extensive experiments on public datasets validate the effectiveness and efficiency of our method. Furthermore, our method runs at 30 fps at the testing time, thus could facilitate many real-time applications such as visual SLAM and human-robot interaction. Code is available at //github.com/svip-lab/PlanarReconstruction.
With the advent of deep neural networks, learning-based approaches for 3D reconstruction have gained popularity. However, unlike for images, in 3D there is no canonical representation which is both computationally and memory efficient yet allows for representing high-resolution geometry of arbitrary topology. Many of the state-of-the-art learning-based 3D reconstruction approaches can hence only represent very coarse 3D geometry or are limited to a restricted domain. In this paper, we propose occupancy networks, a new representation for learning-based 3D reconstruction methods. Occupancy networks implicitly represent the 3D surface as the continuous decision boundary of a deep neural network classifier. In contrast to existing approaches, our representation encodes a description of the 3D output at infinite resolution without excessive memory footprint. We validate that our representation can efficiently encode 3D structure and can be inferred from various kinds of input. Our experiments demonstrate competitive results, both qualitatively and quantitatively, for the challenging tasks of 3D reconstruction from single images, noisy point clouds and coarse discrete voxel grids. We believe that occupancy networks will become a useful tool in a wide variety of learning-based 3D tasks.
We present the first method to capture the 3D total motion of a target person from a monocular view input. Given an image or a monocular video, our method reconstructs the motion from body, face, and fingers represented by a 3D deformable mesh model. We use an efficient representation called 3D Part Orientation Fields (POFs), to encode the 3D orientations of all body parts in the common 2D image space. POFs are predicted by a Fully Convolutional Network (FCN), along with the joint confidence maps. To train our network, we collect a new 3D human motion dataset capturing diverse total body motion of 40 subjects in a multiview system. We leverage a 3D deformable human model to reconstruct total body pose from the CNN outputs by exploiting the pose and shape prior in the model. We also present a texture-based tracking method to obtain temporally coherent motion capture output. We perform thorough quantitative evaluations including comparison with the existing body-specific and hand-specific methods, and performance analysis on camera viewpoint and human pose changes. Finally, we demonstrate the results of our total body motion capture on various challenging in-the-wild videos. Our code and newly collected human motion dataset will be publicly shared.
注冊地址： 北京市海淀區羊坊店路18號2幢3層301-191