اشترك بالحزمة الذهبية واحصل على وصول غير محدود شمرا أكاديميا
تسجيل مستخدم جديدBirds-Eye-View Panoptic Segmentation Using Monocular Frontal View Images
71
0
0.0
(
0
)
نشر من قبل
Nikhil Bharadwaj Gosala
تاريخ النشر
2021
مجال البحث
الهندسة المعلوماتية
والبحث باللغة
English
اسأل ChatGPT حول البحث
ﻻ يوجد ملخص باللغة العربية
Birds-Eye-View (BEV) maps have emerged as one of the most powerful representations for scene understanding due to their ability to provide rich spatial context while being easy to interpret and process. However, generating BEV maps requires complex multi-stage paradigms that encapsulate a series of distinct tasks such as depth estimation, ground plane estimation, and semantic segmentation. These sub-tasks are often learned in a disjoint manner which prevents the model from holistic reasoning and results in erroneous BEV maps. Moreover, existing algorithms only predict the semantics in the BEV space, which limits their use in applications where the notion of object instances is critical. In this work, we present the first end-to-end learning approach for directly predicting dense panoptic segmentation maps in the BEV, given a single monocular image in the frontal view (FV). Our architecture follows the top-down paradigm and incorporates a novel dense transformer module consisting of two distinct transformers that learn to independently map vertical and flat regions in the input image from the FV to the BEV. Additionally, we derive a mathematical formulation for the sensitivity of the FV-BEV transformation which allows us to intelligently weight pixels in the BEV space to account for the varying descriptiveness across the FV image. Extensive evaluations on the KITTI-360 and nuScenes datasets demonstrate that our approach exceeds the state-of-the-art in the PQ metric by 3.61 pp and 4.93 pp respectively.
قيم البحث
اقرأ أيضاً
Recent deep learning models achieve impressive results on 3D scene analysis tasks by operating directly on unstructured point clouds. A lot of progress was made in the field of object classification and semantic segmentation. However, the task of ins
tance segmentation is less explored. In this work, we present 3D-BEVIS, a deep learning framework for 3D semantic instance segmentation on point clouds. Following the idea of previous proposal-free instance segmentation approaches, our model learns a feature embedding and groups the obtained feature space into semantic instances. Current point-based methods scale linearly with the number of points by processing local sub-parts of a scene individually. However, to perform instance segmentation by clustering, globally consistent features are required. Therefore, we propose to combine local point geometry with global context information from an intermediate birds-eye view representation.
Recognizing places using Lidar in large-scale environments is challenging due to the sparse nature of point cloud data. In this paper we present BVMatch, a Lidar-based frame-to-frame place recognition framework, that is capable of estimating 2D relat
ive poses. Based on the assumption that the ground area can be approximated as a plane, we uniformly discretize the ground area into grids and project 3D Lidar scans to birds-eye view (BV) images. We further use a bank of Log-Gabor filters to build a maximum index map (MIM) that encodes the orientation information of the structures in the images. We analyze the orientation characteristics of MIM theoretically and introduce a novel descriptor called birds-eye view feature transform (BVFT). The proposed BVFT is insensitive to rotation and intensity variations of BV images. Leveraging the BVFT descriptors, we unify the Lidar place recognition and pose estimation tasks into the BVMatch framework. The experiments conducted on three large-scale datasets show that BVMatch outperforms the state-of-the-art methods in terms of both recall rate of place recognition and pose estimation accuracy.
Birds-eye-view (BEV) is a powerful and widely adopted representation for road scenes that captures surrounding objects and their spatial locations, along with overall context in the scene. In this work, we focus on birds eye semantic segmentation, a
task that predicts pixel-wise semantic segmentation in BEV from side RGB images. This task is made possible by simulators such as Carla, which allow for cheap data collection, arbitrary camera placements, and supervision in ways otherwise not possible in the real world. There are two main challenges to this task: the view transformation from side view to birds eye view, as well as transfer learning to unseen domains. Existing work transforms between views through fully connected layers and transfer learns via GANs. This suffers from a lack of depth reasoning and performance degradation across domains. Our novel 2-staged perception pipeline explicitly predicts pixel depths and combines them with pixel semantics in an efficient manner, allowing the model to leverage depth information to infer objects spatial locations in the BEV. In addition, we transfer learning by abstracting high-level geometric features and predicting an intermediate representation that is common across different domains. We publish a new dataset called BEVSEG-Carla and show that our approach improves state-of-the-art by 24% mIoU and performs well when transferred to a new domain.
Temporal semantic scene understanding is critical for self-driving cars or robots operating in dynamic environments. In this paper, we propose 4D panoptic LiDAR segmentation to assign a semantic class and a temporally-consistent instance ID to a sequ
ence of 3D points. To this end, we present an approach and a point-centric evaluation metric. Our approach determines a semantic class for every point while modeling object instances as probability distributions in the 4D spatio-temporal domain. We process multiple point clouds in parallel and resolve point-to-instance associations, effectively alleviating the need for explicit temporal data association. Inspired by recent advances in benchmarking of multi-object tracking, we propose to adopt a new evaluation metric that separates the semantic and point-to-instance association aspects of the task. With this work, we aim at paving the road for future developments of temporal LiDAR panoptic perception.
Accurate 3D human pose estimation from single images is possible with sophisticated deep-net architectures that have been trained on very large datasets. However, this still leaves open the problem of capturing motions for which no such database exis
ts. Manual annotation is tedious, slow, and error-prone. In this paper, we propose to replace most of the annotations by the use of multiple views, at training time only. Specifically, we train the system to predict the same pose in all views. Such a consistency constraint is necessary but not sufficient to predict accurate poses. We therefore complement it with a supervised loss aiming to predict the correct pose in a small set of labeled images, and with a regularization term that penalizes drift from initial predictions. Furthermore, we propose a method to estimate camera pose jointly with human pose, which lets us utilize multi-view footage where calibration is difficult, e.g., for pan-tilt or moving handheld cameras. We demonstrate the effectiveness of our approach on established benchmarks, as well as on a new Ski dataset with rotating cameras and expert ski motion, for which annotations are truly hard to obtain.
سجل دخول لتتمكن من نشر تعليقات
التعليقات
جاري جلب التعليقات
سجل دخول لتتمكن من متابعة معايير البحث التي قمت باختيارها
