Filters








59 Hits in 10.4 sec

Reinforced Axial Refinement Network for Monocular 3D Object Detection [article]

Lijie Liu, Chufan Wu, Jiwen Lu, Lingxi Xie, Jie Zhou, Qi Tian
2020 arXiv   pre-print
The proposed framework, Reinforced Axial Refinement Network (RAR-Net), serves as a post-processing stage which can be freely integrated into existing monocular 3D detection methods, and improve the performance  ...  Monocular 3D object detection aims to extract the 3D position and properties of objects from a 2D input image.  ...  In this paper, we ease this challenge by presenting a new framework called Reinforced Axial Refinement Network (RAR-Net), which, as illustrated in Fig. 1 , iteratively refines the detected 3D object to  ... 
arXiv:2008.13748v1 fatcat:5eorvtrt3fgtzmyhetqkxmbr4a

Learning Auxiliary Monocular Contexts Helps Monocular 3D Object Detection [article]

Xianpeng Liu, Nan Xue, Tianfu Wu
2021 arXiv   pre-print
Monocular 3D object detection aims to localize 3D bounding boxes in an input single 2D image.  ...  This paper proposes a simple yet effective formulation for monocular 3D object detection without exploiting any extra information.  ...  Reinforced axial refinement network for monocular 3d ob- Chen, X.; Ma, H.; Wan, J.; Li, B.; and Xia, T. 2017. Multi- ject detection.  ... 
arXiv:2112.04628v1 fatcat:e5ev2xesvjgmpe5cfm57cscu6i

Hybrid Approach for Orientation-Estimation of Rotating Humans in Video Frames Acquired by Stationary Monocular Camera

David Baumgartner, Christoph Praschl, Tobias Zucali, Gerald A. Zwettler
2020 Journal of WSCG  
Thus, novel approaches of Deep Learning for precise object pose-estimation in robotics are hard to adapt to human body analysis.  ...  The human body in-place rotating in T-pose is thereby aligned in the center, applying object tracking methods to compensate for translations of the body movement.  ...  A 3D skeleton can easily be derived if incorporating multiple views with particular 3D approximation to combine the results with a simple neural network [Rho18a] .  ... 
doi:10.24132/csrn.2020.3001.5 fatcat:gqllkqr5gzeqnj3vjcpd5f7msi

Learning an optimal PSF-pair for ultra-dense 3D localization microscopy [article]

Elias Nehme, Boris Ferdman, Lucien E. Weiss, Tal Naor, Daniel Freedman, Tomer Michaeli, Yoav Shechtman
2020 arXiv   pre-print
One established approach for snapshot 3D imaging is point-spread-function (PSF) engineering, in which the PSF is modified to encode the axial information.  ...  Here we suggest using multiple PSFs simultaneously to help overcome this challenge, and investigate the problem of engineering multiple PSFs for dense 3D localization.  ...  Then, the number is increased to 80 for the refinement module.  ... 
arXiv:2009.14303v1 fatcat:naagzp2tfnhgzpjzhp6s7qnwqa

Table of Contents

2019 2019 IEEE/CVF International Conference on Computer Vision (ICCV)  
Wang (Xidian University), Cheng Deng (Xidian University), and Xianglong Liu (Beihang University) Structured Modeling of Joint Deep Feature and Prediction Refinement for Salient Object Detection 3788  ...  Irvine), and Charless Fowlkes (UC Irvine) 3D-RelNet: Joint Object and Relational Network for 3D Prediction 2212 Nilesh Kulkarni (Carnegie Mellon University), Ishan Misra (Facebook AI Research), Shubham  ... 
doi:10.1109/iccv.2019.00004 fatcat:5aouo4scprc75c7zetsimylj2y

Deep Reinforcement Learning in Computer Vision: A Comprehensive Survey [article]

Ngan Le, Vidhiwar Singh Rathour, Kashu Yamazaki, Khoa Luu, Marios Savvides
2021 arXiv   pre-print
(i)landmark localization (ii) object detection; (iii) object tracking; (iv) registration on both 2D image and 3D image volumetric data (v) image segmentation; (vi) videos analysis; and (vii) other applications  ...  Deep reinforcement learning augments the reinforcement learning framework and utilizes the powerful representation of deep neural networks.  ...  To leverage that method, [231] introduced Reinforced Axial Refinement Network (RARN) for monocular 3D object detection by utilizing an RL model to iteratively refining the sampled bounding box to be  ... 
arXiv:2108.11510v1 fatcat:wkkqittwivbx5fpwg3nggcy7cm

A Survey of Deep Learning-based Object Detection

Licheng Jiao, Fan Zhang, Fang Liu, Shuyuan Yang, Lingling Li, Zhixi Feng, Rong Qu
2019 IEEE Access  
With the rapid development of deep learning networks for detection tasks, the performance of object detectors has been greatly improved.  ...  Some representative branches of object detection are analyzed as well.  ...  [295] perform 3D object detection from a single monocular image. Chen et al.  ... 
doi:10.1109/access.2019.2939201 fatcat:jesz2av2tjbkxfpaqyecptgls4

A Platform for Brain-Wide Functional Ultrasound Imaging and Analysis of Circuit Dynamics in Behaving Mice

Clément Brunner, Micheline Grillet, Arnau Sans-Dublanc, Karl Farrow, Théo Lambert, Emilie Mace, Gabriel Montaldo, Alan Urban
2020 Social Science Research Network  
We combined high-frequency 1024-channel 2D-array transducer with advanced multiplexing and highperformance computing for real-time 3D Power Doppler imaging at high spatiotemporal resolution (220x280x175  ...  In addition, we developed a standardized software pipeline for registration and segmentation based on the Allen Mouse Common Coordinate Framework, allowing for temporal analysis in 268 individual brain  ...  Krumin for assistance with the design of the mouse holder and Dr. C. Cowan for help with the Blender software for 3D rendering.  ... 
doi:10.2139/ssrn.3581369 fatcat:wtrbkkfrljbmlpn4z3btgu3hoe

Detecting socially interacting groups using f-formation: A survey of taxonomy, methods, datasets, applications, challenges, and future research directions [article]

Hrishav Bakul Barua, Theint Haythi Mg, Pradip Pramanick, Chayan Sarkar
2021 arXiv   pre-print
The theory of f-formation can be utilized for this purpose. As the types of formations can be very diverse, detecting the social groups is not a trivial task.  ...  In this article, we provide a comprehensive survey of the existing work on social interaction and group detection using f-formation for robotics and other applications.  ...  Secondly, it refines those using a 3D patch trajectory stream and provides temporally stable 3D skeletons. Authors evaluate the method quantitatively and qualitatively yielding an accuracy of 99%.  ... 
arXiv:2108.06181v2 fatcat:walfqfi55fe4fja3imr4qu6asu

Learning Capacity in Simulated Virtual Neurological Procedures

Mattia Samuel Mancosu, Silvester Czanner, Martin Punter
2020 Journal of WSCG  
ACKNOWLEDGEMENTS The authors would like to thank Oana Rotaru-Orhei for her comments and the three anonymous reviewers for their insightful suggestions.  ...  ACKNOWLEDGMENTS The authors acknowledge the support of the NSERC/Creaform Industrial Research Chair on 3-D Scanning for conducting the work presented in this paper.  ...  Such applications require object detection capabilities for monitoring based on edge detection procedures very frequently supported in the outcomes of the application of filters for edge detection, which  ... 
doi:10.24132/csrn.2020.3001.13 fatcat:uytlm7nytrhmnk553ellfhl54a

A platform for brain-wide functional ultrasound imaging and analysis of circuit dynamics in behaving mice [article]

Clément BRUNNER, Micheline GRILLET, Arnau SANS-DUBLANC, Karl FARROW, Théo LAMBERT, Émilie MACÉ, Gabriel MONTALDO, Alan URBAN
2020 bioRxiv   pre-print
We combined high-frequency 1024-channel 2D-array transducer with advanced multiplexing and high-performance computing for real-time 3D Power Doppler imaging at high spatiotemporal resolution (220x280x175  ...  In addition, we developed a standardized software pipeline for registration and segmentation based on the Allen Mouse Common Coordinate Framework, allowing for temporal analysis in 268 individual brain  ...  Krumin for assistance with the design of the mouse holder and Dr. C. Cowan for help with the Blender software for 3D rendering.  ... 
doi:10.1101/2020.04.10.035436 fatcat:efjwtxhdevbs3j24qlkmrz7inq

Application and Construction of Deep Learning Networks in Medical Imaging

Maribel Torres-Velazquez, Wei-Jie Chen, Xue Li, Alan B. McMillan
2020 IEEE Transactions on Radiation and Plasma Medical Sciences  
As a result of this advantage, DL has been successfully applied within the medical imaging field to address problems, such as disease classification and tumor segmentation for which it is difficult or  ...  The sections of this review summarize the milestones related to the development of the DL field, followed by a description of the elements of deep neural network and an overview of its application within  ...  The objective of classification is to sort input images into two or more discrete classes, such as to detect the presence or absence of disease.  ... 
doi:10.1109/trpms.2020.3030611 pmid:34017931 pmcid:PMC8132932 fatcat:tp7yk5wqwbga5imte7un533vci

27th Annual Computational Neuroscience Meeting (CNS*2018): Part Two

2018 BMC Neuroscience  
Dynamic updating of working memory resources for visual objects.  ...  The striatum is a structure of the basal ganglia that is critical for reinforcement learning.  ...  Our results suggest that MLMC may offer significant speed-up for collecting statistics from spiking network models, particularly for predominantly feed-forward networks and for recurrent networks operating  ... 
doi:10.1186/s12868-018-0451-y fatcat:afgrjlnjgjarldkuwo3e2pt5sm

A neural model of 3D shape-from-texture: Multiple-scale filtering, boundary grouping, and surface filling-in

Stephen Grossberg, Levin Kuhlmann, Ennio Mingolla
2007 Vision Research  
In particular, the model represents a high degree of 3D curvature for a certain class of images, all of whose texture elements have the same degree of optical compression, in accordance with percepts of  ...  This is achieved in the model through multiple-scale filtering of a 2D image, followed by a cooperative-competitive grouping network that coherently binds texture elements into boundary webs at the appropriate  ...  Across-depth competition helps to refine the 3D boundary representations.  ... 
doi:10.1016/j.visres.2006.10.024 pmid:17275061 fatcat:46i5ttl2x5a27fxiij45o52hwy

Conference Digest

2020 2020 IEEE Aerospace Conference  
using innovative techniques in Radar Data Processing (RDP) for a 3D radar having Primary and Secondary detection.  ...  for independent verification necessitated robust methods of monocular localization.  ...  Developed for NASA Ames Research Center, ADTM leverages proven artificial intelligence techniques for rapidly detecting and diagnosing anomalies in near real-time.  ... 
doi:10.1109/aero47225.2020.9172613 fatcat:ioqf5ijrx5gvffu3ls34aa2nsq
« Previous Showing results 1 — 15 out of 59 results