3,785 Hits in 3.5 sec

Depth-Aware CNN for RGB-D Segmentation [chapter]

Weiyue Wang, Ulrich Neumann
2018 Lecture Notes in Computer Science  
The availability of depth data enables progress in RGB-D semantic segmentation with CNNs.  ...  To address these issues, we present Depth-aware CNN by introducing two intuitive, flexible and effective operations: depth-aware convolution and depth-aware average pooling.  ...  Acknowledgements We thank Ronald Yu, Yi Zhou and Qiangui Huang for the discussion and proofread.  ... 
doi:10.1007/978-3-030-01252-6_9 fatcat:ecvunnbzqvb3xj7xk7sbmaxgmu

Depth-aware CNN for RGB-D Segmentation [article]

Weiyue Wang, Ulrich Neumann
2018 arXiv   pre-print
The availability of depth data enables progress in RGB-D semantic segmentation with CNNs.  ...  To address these issues, we present Depth-aware CNN by introducing two intuitive, flexible and effective operations: depth-aware convolution and depth-aware average pooling.  ...  Acknowledgements We thank Ronald Yu, Yi Zhou and Qiangui Huang for the discussion and proofread.  ... 
arXiv:1803.06791v1 fatcat:dxogeqi4grc6ngmxiwpvmbjk5a

Applying Surface Normal Information in Drivable Area and Road Anomaly Detection for Ground Mobile Robots [article]

Hengli Wang, Rui Fan, Yuxiang Sun, Ming Liu
2020 arXiv   pre-print
Our NIM can be deployed in existing convolutional neural networks (CNNs) to refine the segmentation performance.  ...  In recent years, many impressive semantic segmentation networks, which can be used for pixel-level drivable area and road anomaly detection, have been developed.  ...  (6) NIM-RGB+D (Ours), and (j) Depth-aware CNN and (k) RTFNet.  ... 
arXiv:2008.11383v1 fatcat:pukh4rfnd5hz3mizhh7ibccqmq

High-quality Instance-aware Semantic 3D Map Using RGB-D Camera [article]

Dinh-Cuong Hoang, Todor Stoyanov, Achim J. Lilienthal
2019 arXiv   pre-print
In this work, we integrate deep-learning-based instance segmentation and classification into a state of the art RGB-D SLAM system.  ...  We present a mapping system capable of constructing detailed instance-level semantic models of room-sized indoor environments by means of an RGB-D camera.  ...  The image space domain is defined as Ω ⊂ N 2 , where an RGB-D frame is composed of a color map and a depth map D of depth pixels d : Ω → R.  ... 
arXiv:1903.10782v6 fatcat:v6ncwi2b5jhjnnldaenxieromq

Depth-Adapted CNNs for RGB-D Semantic Segmentation [article]

Zongwei Wu, Guillaume Allibert, Christophe Stolz, Chao Ma, Cédric Demonceaux
2022 arXiv   pre-print
Recent RGB-D semantic segmentation has motivated research interest thanks to the accessibility of complementary modalities from the input side.  ...  In this paper, we propose a novel framework to incorporate the depth information in the RGB convolutional neural network (CNN), termed Z-ACN (Depth-Adapted CNN).  ...  Inspired by these works, we propose to compute the non-local awareness from the depth priors, making the convolution geometry-aware for RGB-D semantic segmentation.  ... 
arXiv:2206.03939v1 fatcat:pvsj5euasbdcdfb4aqckhot2lm

Geometry-Aware Distillation for Indoor Semantic Segmentation

Jianbo Jiao, Yunchao Wei, Zequn Jie, Honghui Shi, Rynson Lau, Thomas S. Huang
2019 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)  
It has been shown that jointly reasoning the 2D appearance and 3D information from RGB-D domains is beneficial to indoor scene semantic segmentation.  ...  In this paper, we propose to jointly infer the semantic and depth information by distilling geometry-aware embedding to eliminate such strong constraint while still exploiting the helpful depth domain  ...  Similar to recent RGB semantic segmentation, CNN also benefits RGB-D approaches.  ... 
doi:10.1109/cvpr.2019.00298 dblp:conf/cvpr/JiaoWJSLH19 fatcat:u4pkeaiatbaktp2ihwoigzkjgq

Self-Supervised Drivable Area and Road Anomaly Segmentation using RGB-D Data for Robotic Wheelchairs [article]

Hengli Wang, Yuxiang Sun, Ming Liu
2020 arXiv   pre-print
Then, we train RGB-D data-based semantic segmentation neural networks and get predicted labels.  ...  We develop a pipeline that can automatically generate segmentation labels for drivable areas and road anomalies.  ...  Here, we use three off-the-shelf RGB-D data-based semantic segmentation neural networks, FuseNet [5] , Depth-aware CNN [6] and RTFNet [28] .  ... 
arXiv:2007.05950v1 fatcat:75u4tzt5szccnbmmym4qv55u7q

Variational Context-Deformable ConvNets for Indoor Scene Parsing

Zhitong Xiong, Yuan Yuan, Nianhui Guo, Qi Wang
2020 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)  
the training of VCD module, which can make it continuous and more stable; 3) a perspective-aware guidance module is designed to take advantage of multi-modal information for RGB-D segmentation.  ...  Context information is critical for image semantic segmentation.  ...  Towards perspective understanding of RGB-D scene, Depth-aware CNNs [47] built a depth-aware receptive field to augment standard convolution.  ... 
doi:10.1109/cvpr42600.2020.00405 dblp:conf/cvpr/Xiong0G020 fatcat:ijodulywevgandvwemvkznyt6a

ShapeConv: Shape-aware Convolutional Layer for Indoor RGB-D Semantic Segmentation [article]

Jinming Cao, Hanchao Leng, Dani Lischinski, Danny Cohen-Or, Changhe Tu, Yangyan Li
2021 arXiv   pre-print
RGB-D semantic segmentation has attracted increasing attention over the past few years.  ...  Extensive experiments on three challenging indoor RGB-D semantic segmentation benchmarks, i.e., NYU-Dv2(-13,-40), SUN RGB-D, and SID, demonstrate the effectiveness of our ShapeConv when employing it over  ...  This simple replacement transforms CNNs designed for RGB data into ones better suited for consuming RGB-D data.  ... 
arXiv:2108.10528v1 fatcat:or2zcvlaubgdjjbwe4rhjzzxfe

Boundary-aware CNN for Semantic Segmentation

Nan Zou, Zhiyu Xiang, Yiman Chen, Shuya Chen, Chengyu Qiao
2019 IEEE Access  
The whole network is implemented end-to-end and evaluated with heterogeneous RGB and depth input.  ...  Experiments conducted on multiple datasets show that our boundary-aware CNN can effectively improve the semantic segmentation performance.  ...  To well utilize the depth information, depth-aware CNN [8] was proposed.  ... 
doi:10.1109/access.2019.2935816 fatcat:2bvsdqrr2be47bfvbbqcoixkdy

Modality and Component Aware Feature Fusion for RGB-D Scene Classification

Anran Wang, Jianfei Cai, Jiwen Lu, Tat-Jen Cham
2016 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR)  
The CNN features are computed from an augmented pixel-wise representation comprising multiple modalities of RGB, HHA and surface normals, as extracted from RGB-D data.  ...  Depth Dataset V2.  ...  [7, 6] described a method to detect contours in RGB-D images and use them for semantic segmentation, further treating the quantized semantic segmentation output as local features for scene classification  ... 
doi:10.1109/cvpr.2016.645 dblp:conf/cvpr/0001CLC16 fatcat:kb3tcjeb5ngyrexmptrxcupxgq

Malleable 2.5D Convolution: Learning Receptive Fields along the Depth-axis for RGB-D Scene Parsing [article]

Yajie Xing, Jingbo Wang, Gang Zeng
2020 arXiv   pre-print
Depth data provide geometric information that can bring progress in RGB-D scene parsing tasks.  ...  We conduct extensive experiments on two challenging RGB-D semantic segmentation dataset NYUDv2 and Cityscapes to validate the effectiveness and the generalization ability of our method.  ...  Wang, W., Neumann, U.: Depth-aware CNN for RGB-D segmentation. In: ECCV (11). Lecture Notes in Computer Science, vol. 11215, pp. 144-161. Springer (2018) 32.  ... 
arXiv:2007.09365v1 fatcat:iyaz3haldrfpphx6365o7ebhd4

Deep Learning-Based 3D Instance and Semantic Segmentation: A Review

Siddiqui Muhammad Yasir, Hyunsik Ahn
2022 Journal on Artificial Intelligence  
The process of segmenting point cloud data into several homogeneous areas with points in the same region having the same attributes is known as 3D segmentation.  ...  This study examines many strategies that have been presented to 3D instance and semantic segmentation and gives a complete assessment of current developments in deep learning-based 3D segmentation.  ...  [40] presented a depth-aware 2D-CNN by incorporating two unique layers, a depth aware convolution layer and a depth-aware pooling layer, based on the assumption that pixels with the same semantic label  ... 
doi:10.32604/jai.2022.031235 fatcat:nsxfjoy4mnamfkrsu2mgsoums4

Learning Depth-Aware Deep Representations for Robotic Perception

Lorenzo Porzi, Samuel Rota Bulo, Adrian Penate-Sanchez, Elisa Ricci, Francesc Moreno-Noguer
2017 IEEE Robotics and Automation Letters  
Most existing approaches, however, exploit RGB-D data by simply considering depth as an additional input channel for the network.  ...  Exploiting RGB-D data by means of Convolutional Neural Networks (CNNs) is at the core of a number of robotics applications, including object detection, scene semantic segmentation and grasping.  ...  RELATED WORK CNNs for RGB-D data.  ... 
doi:10.1109/lra.2016.2637444 dblp:journals/ral/PorziBSRM17 fatcat:h2m2dok56jdxlmxcs4ja25mkvm

Effective Free-Driving Region Detection for Mobile Robots by Uncertainty Estimation Using RGB-D Data

Toan-Khoa Nguyen, Phuc Thanh-Thien Nguyen, Dai-Dong Nguyen, Chung-Hsien Kuo
2022 Sensors  
Then, we train RGB-D datasets with a semantic segmentation network using self-generated ground truth labels derived from our method (AGSL labels) to get the pre-trained model.  ...  First, we propose the Automatic Generating Segmentation Label (AGSL) framework, which is an efficient system automatically generating segmentation labels for drivable areas and road anomalies by finding  ...  , Depth-aware CNN, and RTFNet, respectively.  ... 
doi:10.3390/s22134751 pmid:35808244 pmcid:PMC9268933 fatcat:mrzqz6kms5hprlghy7jqyo67di
« Previous Showing results 1 — 15 out of 3,785 results