Filters








7,189 Hits in 7.1 sec

Monocular 3D Object Detection with Depth from Motion [article]

Tai Wang, Jiangmiao Pang, Dahua Lin
2022 arXiv   pre-print
Perceiving 3D objects from monocular inputs is crucial for robotic systems, given its economy compared to multi-sensor settings.  ...  Motivated by binocular methods for 3D object detection, we take advantage of the strong geometry structure provided by camera ego-motion for accurate object depth estimation and detection.  ...  Learningbased video depth estimation methods can be divided into MVS-based (Multi-View-Stereo) methods [23, 42] and monocular-stereo hybrid methods [53, 28, 18] .  ... 
arXiv:2207.12988v1 fatcat:rj2hphm4vzaljb5xc2elfwujku

Stereo-Based vs. Monocular 6-DoF Pose Estimation Using Point Features: A Quantitative Comparison [chapter]

Pedram Azad, Tamim Asfour, Rüdiger Dillmann
2009 Autonome Mobile Systeme 2009  
In the recent past, object recognition and localization based on correspondences of local point features in 2D views has become very popular in the robotics community.  ...  For grasping and manipulation with robotic systems, in addition accurate 6-DoF pose estimation of the object of interest is necessary.  ...  Such methods all have in common that the full pose of the object is computed on the basis of a monocular image.  ... 
doi:10.1007/978-3-642-10284-4_6 dblp:conf/ams/AzadAD09 fatcat:iscrtz5ycbgwddbyadciqfkaha

A novel fractal monocular and stereo video codec with object-based functionality

Shiping Zhu, Liyun Li, Zaikuo Wang
2012 EURASIP Journal on Advances in Signal Processing  
The proposed object-based fractal monocular and stereo video coding methods are simple and effective, and they make the applications of fractal monocular and stereo video coding more flexible and practicable  ...  The PSNR of the proposed stereo video coding is about 0.17 dB higher than that of the proposed monocular video coding, and 0.69 dB higher than that of JMVC 4.0 on average.  ...  Overseas Chinese Scholars from the State Education Ministry of China.  ... 
doi:10.1186/1687-6180-2012-227 fatcat:z4gzoc3f6jdkxknxkhz6eu3q34

Pseudo-Stereo for Monocular 3D Object Detection in Autonomous Driving [article]

Yi-Nan Chen and Hang Dai and Yong Ding
2022 arXiv   pre-print
The advanced stereo 3D detectors can also accurately localize 3D objects. The gap in image-to-image generation for stereo views is much smaller than that in image-to-LiDAR generation.  ...  detecting 3D objects from a single image.  ...  Then, we feed the Pseudo-Stereo views to stereo 3D detectors for detecting 3D objects from the single input image.  ... 
arXiv:2203.02112v1 fatcat:z6z2ex3dkfegtgfk4xn7qmmj4y

3D Object Detection for Autonomous Driving: A Review and New Outlooks [article]

Jiageng Mao, Shaoshuai Shi, Xiaogang Wang, Hongsheng Li
2022 arXiv   pre-print
and planning. 3D object detection, which intelligently predicts the locations, sizes, and categories of the critical 3D objects near an autonomous vehicle, is an important part of a perception system.  ...  Additionally, we systematically investigate the applications of 3D object detection in driving systems.  ...  View Meta/Graph Kernel Convolution Range Image Range View (RV) 3D Detection from Bird's-Eye View Transform into Bird's Eye View (BEV) Range Image Range View (RV) 3D Detection from Point View Transform  ... 
arXiv:2206.09474v1 fatcat:3skws77uqngjtpo6mycpo4dhny

Visually Imbalanced Stereo Matching

Yicun Liu, Jimmy Ren, Jiawei Zhang, Jianbo Liu, Mude Lin
2020 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)  
We show that resembling the human visual system, those algorithms can handle limited degrees of monocular downgrading but also prone to collapses beyond a certain threshold.  ...  In this paper, we carry out a systematic comparison to investigate the effect of various imbalanced conditions on current popular stereo matching algorithms.  ...  Instead of predicting the image after spatial transformation, it is more efficient to estimate the transformation matrix itself.  ... 
doi:10.1109/cvpr42600.2020.00210 dblp:conf/cvpr/LiuRZLL20 fatcat:2op5thaezzddvhz6f3wjzdehsa

A Novel Monocular Disparity Estimation Network with Domain Transformation and Ambiguity Learning [article]

Juan Luis Gonzalez Bello, Munchurl Kim
2019 arXiv   pre-print
Recent works have uncovered the advantages of using an unsupervised scheme to train CNN's to estimate monocular disparity, where only the relatively-easy-to-obtain stereo images are needed for training  ...  Convolutional neural networks (CNN) have shown state-of-the-art results for low-level computer vision problems such as stereo and monocular disparity estimations, but still, have much room to further improve  ...  INTRODUCTION For a given object displayed in a rectified stereo pair of images, the disparity is defined as the horizontal pixel distance of the object in the left and right images.  ... 
arXiv:1903.08514v1 fatcat:bx3y7jkhmbgnjdk7xdf3eti7nu

Stereo-based simultaneous localization, mapping and moving object tracking

Kuen-Han Lin, Chieh-Chih Wang
2010 2010 IEEE/RSJ International Conference on Intelligent Robots and Systems  
However, monocular SLAMMOT inherits the observability issue of bearings-only tracking in which moving entities would be unobservable according to motions of the camera and moving objects.  ...  Simulation and experimental results demonstrate that the proposed stereo SLAMMOT is superior than monocular SLAMMOT in dynamic environments.  ...  (a)(b)(c) showed the stereo image pairs and the results at frame 136, 315 and 468. The upper image is the right camera view and the lower image is the left camera view.  ... 
doi:10.1109/iros.2010.5649653 dblp:conf/iros/LinW10 fatcat:j3f4tqticjfplafx3fw7jjy76y

Multi-Sensor 3D Object Box Refinement for Autonomous Driving [article]

Peiliang Li, Siqi Liu, Shaojie Shen
2019 arXiv   pre-print
We demonstrate superior 3D detection and localization performance compared to state-of-the-art monocular, stereo methods and competitive performance compared with the baseline LiDAR method on the KITTI  ...  ., pixels for stereo, 3D points for LiDAR), we model the local geometry as an instance vector representation, which indicates the 3D coordinate of each element respecting to the object frame.  ...  [26] , [24] , [27] encode the point cloud into 2D image formats by projecting raw 3D points to front view or bird's eye view images, the 3D object detection can thereby be achieved using the regular  ... 
arXiv:1909.04942v2 fatcat:pf3ezflxdbg3lehhirh5zzskia

Beyond Monocular Deraining: Parallel Stereo Deraining Network Via Semantic Prior [article]

Kaihao Zhang, Wenhan Luo, Yanjiang Yu, Wenqi Ren, Fang Zhao, Changsheng Li, Lin Ma, Wei Liu, Hongdong Li
2021 arXiv   pre-print
Experiments on both monocular and the newly proposed stereo rainy datasets demonstrate that the proposed method achieves the state-of-the-art performance.  ...  Taking images in the rain however often results in degraded quality of images, thus compromises the performance of many computer vision systems.  ...  Acknowledgment This work is funded in part by the ARC Centre of Excellence for Robotics Vision (CE140100016), ARC-Discovery (DP 190102261) and ARC-LIEF (190100080) grants.  ... 
arXiv:2105.03830v1 fatcat:icmkotgvsbbpppoxywuykyqbrq

MV-FCOS3D++: Multi-View Camera-Only 4D Object Detection with Pretrained Monocular Backbones [article]

Tai Wang, Qing Lian, Chenming Zhu, Xinge Zhu, Wenwei Zhang
2022 arXiv   pre-print
It is built upon a simple monocular detector FCOS3D++, pretrained only with object annotations of Waymo, and converts multi-view features to a 3D grid space to detect 3D objects thereon.  ...  For multi-view camera-only 3D detection, methods based on bird-eye-view or 3D geometric representations can leverage the stereo cues from overlapped regions between adjacent views and directly perform  ...  of image.  ... 
arXiv:2207.12716v1 fatcat:brhxt72c5fbh3gdzfggpgn5y3i

Single View Stereo Matching

Yue Luo, Jimmy Ren, Mude Lin, Jiahao Pang, Wenxiu Sun, Hongsheng Li, Liang Lin
2018 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition  
The model also generalizes well to other monocular depth estimation benchmarks. We also discuss the implications and the advantages of solving monocular depth estimation using stereo methods. 1  ...  In this paper, we show for the first time that the monocular depth estimation problem can be reformulated as two sub-problems, a view synthesis procedure followed by stereo matching, with two intriguing  ...  View Synthesis Network Stereo Matching Network Left image Synthesized right image Predicted depth Figure 1 : Pipeline of our approach on monocular depth estimation.  ... 
doi:10.1109/cvpr.2018.00024 dblp:conf/cvpr/LuoRLPSLL18 fatcat:5cm6g3ouljgl7cpmx2k6uwic5y

Single View Stereo Matching [article]

Yue Luo, Jimmy Ren, Mude Lin, Jiahao Pang, Wenxiu Sun, Hongsheng Li, Liang Lin
2018 arXiv   pre-print
The model also generalizes well to other monocular depth estimation benchmarks. We also discuss the implications and the advantages of solving monocular depth estimation using stereo methods.  ...  In this paper, we show for the first time that the monocular depth estimation problem can be reformulated as two sub-problems, a view synthesis procedure followed by stereo matching, with two intriguing  ...  View Synthesis Network Stereo Matching Network Left image Synthesized right image Predicted depth Figure 1 : Pipeline of our approach on monocular depth estimation.  ... 
arXiv:1803.02612v2 fatcat:m6f2suvyqfcnrexifyyif3df3q

Eliminating the Blind Spot: Adapting 3D Object Detection and Monocular Depth Estimation to 360$$^\circ $$∘ Panoramic Imagery [chapter]

Grégoire Payen de La Garanderie, Amir Atapour Abarghouei, Toby P. Breckon
2018 Lecture Notes in Computer Science  
To address the lack of annotated panoramic automotive datasets availability, we adapt contemporary automotive dataset, via style and projection transformations, to facilitate the cross-domain retraining  ...  Following this approach we retrain and adapt existing architectures to recover scene depth and 3D pose of vehicles from monocular panoramic imagery without any panoramic training labels or calibration  ...  The rectilinear projection as defined in Eqn. 1 is advantageous because the camera matrix P can be combined with further image and object space transformations into a single linear transformation followed  ... 
doi:10.1007/978-3-030-01261-8_48 fatcat:yy27el2drbgm5fjfusbxbviimq

Eliminating the Blind Spot: Adapting 3D Object Detection and Monocular Depth Estimation to 360 Panoramic Imagery [article]

Grégoire Payen de La Garanderie, Amir Atapour Abarghouei, Toby P. Breckon
2018 arXiv   pre-print
Following this approach we retrain and adapt existing architectures to recover scene depth and 3D pose of vehicles from monocular panoramic imagery without any panoramic training labels or calibration  ...  To address the lack of annotated panoramic automotive datasets availability, we adapt a contemporary automotive dataset, via style and projection transformations, to facilitate the cross-domain retraining  ...  The rectilinear projection as defined in Eqn. 1 is advantageous because the camera matrix P can be combined with further image and object space transformations into a single linear transformation followed  ... 
arXiv:1808.06253v1 fatcat:uwz4ajy3rjdvfg5djkngyoouvq
« Previous Showing results 1 — 15 out of 7,189 results