Filters








110,296 Hits in 5.2 sec

Bringing Generalization to Deep Multi-View Pedestrian Detection [article]

Jeet Vora, Swetanjal Dutta, Kanishk Jain, Shyamgopal Karthik, Vineet Gandhi
2022 arXiv   pre-print
Multi-view Detection (MVD) is highly effective for occlusion reasoning in a crowded environment.  ...  properties essential to bring generalization to MVD and propose a barebones model to incorporate them.  ...  Some example frames from the proposed Generalized Multi-View Detection (GMVD) dataset are illustrated in Figure 3 .  ... 
arXiv:2109.12227v4 fatcat:26umtpyx2nfe5jb6hzecy3jhfm

RePre: Improving Self-Supervised Vision Transformer with Reconstructive Pre-training [article]

Luya Wang, Feng Liang, Yangguang Li, Honggang Zhang, Wanli Ouyang, Jing Shao
2022 arXiv   pre-print
The multi-hierarchy features provide rich supervisions from low to high semantic information, which are crucial for our RePre.  ...  Our RePre brings decent improvements on various contrastive frameworks with different vision transformer architectures.  ...  We get low-to-high semantic (multi-hierarchy) features from the transformer encoder by sampling shallow-to-deep transformer blocks.  ... 
arXiv:2201.06857v2 fatcat:alqu2db7xngmphbc23w7gwoobi

MVP-Net: Multi-view FPN with Position-aware Attention for Deep Universal Lesion Detection [article]

Zihao Li, Shu Zhang, Junge Zhang, Kaiqi Huang, Yizhou Wang, Yizhou Yu
2019 arXiv   pre-print
Specifically, as radiologists tend to inspect multiple windows for an accurate diagnosis, we explicitly model this process and propose a multi-view feature pyramid network (FPN), where multi-view features  ...  are extracted from images rendered with varied window widths and window levels; to effectively combine this multi-view information, we further propose a position-aware attention module.  ...  We would like to thank Feng Liu for valuable discussions.  ... 
arXiv:1909.04247v3 fatcat:d2g3xgxbpffsbnxsnjriu7byoi

Self-Supervised Multisensor Change Detection [article]

Sudipan Saha, Patrick Ebel, Xiao Xiang Zhu
2022 arXiv   pre-print
Such constraints limit the scope of traditional supervised machine learning and unsupervised generative approaches for multi-sensor change detection.  ...  Adding to this, change detection methods are often constrained to use only target image-pair, no labeled data, and no additional unlabeled data.  ...  Without loss of generality, we henceforth explain the deep clustering process in reference to a generic pixel y b 1The dimension of y b 1,n is K that can be converted to 1-dimensional label c b 1,n by  ... 
arXiv:2103.05102v3 fatcat:p5ba7pq7i5hxrlkwgqoatk7fwe

A Comprehensive Survey on Community Detection with Deep Learning [article]

Xing Su, Shan Xue, Fanzhen Liu, Jia Wu, Jian Yang, Chuan Zhou, Wenbin Hu, Cecile Paris, Surya Nepal, Di Jin, Quan Z. Sheng, Philip S. Yu
2021 arXiv   pre-print
We then discuss the practical applications of community detection in various domains and point to implementation scenarios.  ...  The main category, i.e., deep neural networks, is further divided into convolutional networks, graph attention networks, generative adversarial networks and autoencoders.  ...  Deep learning models [60] bring the following additional advantages to community detection. Thus, deep learning-based community detection has been a new emerging branch.  ... 
arXiv:2105.12584v2 fatcat:matipshxnzcdloygrcrwx2sxr4

Improved computer-aided detection of pulmonary nodules via deep learning in the sinogram domain

Yongfeng Gao, Jiaxing Tan, Zhengrong Liang, Lihong Li, Yumei Huo
2019 Visual Computing for Industry, Biomedicine, and Art  
This study indicates that pulmonary nodule detection in the sinogram domain is feasible with deep learning.  ...  Moreover, a combination of sinogram and CT image could further improve the value of AUC to 0.92.  ...  Comparison of different ways of setting multi-channel inputs From the above experiments, we can see the more projection view can bring us more information for achieving a high performance.  ... 
doi:10.1186/s42492-019-0029-2 pmid:32240409 fatcat:naby34bwszculk6mzc3kv3tyey

Reconstructing piecewise planar scenes with multi-view regularization

Weijie Xi, Xuejin Chen
2019 Computational Visual Media  
Recent approaches for single-view reconstruction employ multi-branch neural networks to simultaneously segment planes and recover 3D plane parameters.  ...  Our multi-view regularization enables the consistency among multiple views by making the feature embedding more robust against view change and lighting variations.  ...  Recently, deep learning techniques widely success in image recognition and detection, and many deep learningbased methods [3] [4] [5] [6] have been proposed for 3D reconstruction.  ... 
doi:10.1007/s41095-019-0159-7 fatcat:kftb6h2lefcg7mlh3zpizc4xhm

Convolutional Neural Networks for Aerial Multi-Label Pedestrian Detection [article]

Amir Soleimani, Nasser M. Nasrabadi
2018 arXiv   pre-print
First, a deep object detector, Single Shot Multibox Detector (SSD), is used to generate object proposals from small aerial images.  ...  The low resolution of objects of interest in aerial images makes pedestrian detection and action detection extremely challenging tasks.  ...  Among the deep detectors, SSD has outperformed other methods, because it detects objects in a multi-scale framework.  ... 
arXiv:1807.05983v1 fatcat:z5nman7hbbagnjohqkvejwbwvq

Pedestrian Detection with Multi-View Convolution Fusion Algorithm

Yuhong Liu, Chunyan Han, Lin Zhang, Xin Gao
2022 Entropy  
With the introduction of the multi-view method, the task of pedestrian detection in crowded or fuzzy scenes has been significantly improved and has become a widely used method in autonomous driving.  ...  In recent years, the pedestrian detection technology of a single 2D image has been dramatically improved.  ...  A multi-view aggregation approach with feature projection allows the generated bird's eye view to aggregate multi-camera pedestrian information.  ... 
doi:10.3390/e24020165 pmid:35205460 pmcid:PMC8870950 fatcat:wff7suvrcfdbzlqbsidtskrv3y

Weakly Supervised Deep Learning for Thoracic Disease Classification and Localization on Chest X-rays

Chaochao Yan, Jiawen Yao, Ruoyu Li, Zheng Xu, Junzhou Huang
2018 Proceedings of the 2018 ACM International Conference on Bioinformatics, Computational Biology, and Health Informatics - BCB '18  
In this article, we propose a weakly supervised deep learning framework equipped with squeeze-and-excitation blocks, multi-map transfer, and max-min pooling for classifying thoracic diseases as well as  ...  While detecting thoracic diseases on chest X-rays is still a challenging task for machine intelligence, due to 1) the highly varied appearance of lesion areas on X-rays from patients of different thoracic  ...  Consequently, a multi-view version of chest X-ray dataset -MIMIC-CXR was presented in [26] , and based on which a dual deep convolutional network framework was naturally proposed to utilize both frontal  ... 
doi:10.1145/3233547.3233573 dblp:conf/bcb/YanYL0H18 fatcat:rjwwuckkz5dktj6u6mct73lfhi

Secure IoT Systems Monitor Framework using Probabilistic Image Encryption

P.G. Narmadha, K. Pooja, K . Muthulakshmi
2020 International Journal of Advanced engineering Management and Science  
The main purpose of this survey is to identify existing methods extensively, and to characterize the literature in a manner that brings to attention key challenges.  ...  Many systems include detection, storage of video information, and human-computer interfaces.  ...  Methods that attempt to derive description from single-view videos do not produce an appropriate set of members when summing up multi-view videos. A. A.  ... 
doi:10.22161/ijaems.66.6 fatcat:q7cxfx4775duveit4cjqk5wgva

On the Fusion of Text Detection Results: A Genetic Programming Approach

Jose L. Flores Campana, Allan Pinto, Manuel Cordova Neira, Luis G. L. Decker, Andreza Santos, Jhonatas S. Conceicao, Ricardo da Silva Torres
2020 IEEE Access  
Another issue refers to the lack of the use of the complementary views provided by different text detection methods. This paper aims to fill these gaps.  ...  INDEX TERMS Scene text detection, multi-oriented text, convolutional neural network, data fusion, genetic programming.  ...  ACKNOWLEDGMENT The authors would like to thank Samsung R&D Institute, Brazil.  ... 
doi:10.1109/access.2020.2987869 fatcat:fcxoeuvonbdjtlvvzpclvs22de

The Elements of End-to-end Deep Face Recognition: A Survey of Recent Advances [article]

Hang Du, Hailin Shi, Dan Zeng, Xiao-Ping Zhang, Tao Mei
2021 arXiv   pre-print
Then, the face alignment is proceeded to calibrate the faces to the canonical view and crop them with a normalized pixel size.  ...  To achieve this, a typical end-to-end system is built with three key elements: face detection, face alignment, and face representation. The face detection locates faces in the image or frame.  ...  Moreover, certain methods [1, 103, 154] developed multiple pose-specific deep models to compute the multi-view face representations. Racial bias: Racial bias is another issue in face recognition.  ... 
arXiv:2009.13290v4 fatcat:vlconzbbyzee5g3s7xnbjgv3ey

SA-NET.V2: REAL-TIME VEHICLE DETECTION FROM OBLIQUE UAV IMAGES WITH USE OF UNCERTAINTY ESTIMATION IN DEEP META-LEARNING

M. Khoshboresh-Masouleh, R. Shah-Hosseini
2022 The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences  
Real-time vehicle detection is more difficult due to the variety of depth and scale vehicles in oblique view UAV images.  ...  a small training dataset, while many vehicle monitoring approaches appear to understand single-time detection with a big training dataset.  ...  Deep meta-learning is an inductive transfer system whose main goal is to improve generalization ability for multiple tasks (Huisman, van Rijn, and Plaat 2021) .  ... 
doi:10.5194/isprs-archives-xlvi-m-2-2022-141-2022 fatcat:xaggev7cvnetfa2bg4h5yhat5y

EM-NET: Centerline-Aware Mitochondria Segmentation in EM Images via Hierarchical View-Ensemble Convolutional Network [article]

Zhimin Yuan, Jiajin Yi, Zhengrong Luo, Zhongdao Jia, Jialin Peng
2020 arXiv   pre-print
To achieve a light-weight 3D network, we introduce a novel hierarchical view-ensemble convolution module to reduce number of parameters, and facilitate multi-view information aggregation.Validations on  ...  To address these problems, we introduce a multi-task network named EM-Net, which includes an auxiliary centerline detection task to account for shape information of mitochondria represented by centerline  ...  In this way, multi-scale and long-range multi-view context information, which are critical to the complicated EM image segmentation, can be encoded in a single module with significantly reduced parameters  ... 
arXiv:1912.00201v3 fatcat:o6r3a32jdbeojeslks27aclmaq
« Previous Showing results 1 — 15 out of 110,296 results