Filters








9,406 Hits in 7.2 sec

Corner Proposal Network for Anchor-free, Two-stage Object Detection [article]

Kaiwen Duan, Lingxi Xie, Honggang Qi, Song Bai, Qingming Huang, Qi Tian
2020 arXiv   pre-print
This paper proposes a novel anchor-free, two-stage framework which first extracts a number of object proposals by finding potential corner keypoint combinations and then assigns a class label to each proposal  ...  Our approach, dubbed Corner Proposal Network (CPN), enjoys the ability to detect objects of various scales and also avoids being confused by a large number of false-positive proposals.  ...  We focus on two important choices of object detection, namely, whether to use anchor-based or anchor-free methods for proposal extraction, and whether to use one-stage or two-stage methods for determining  ... 
arXiv:2007.13816v1 fatcat:6ad34mue6rb37gxfgu5bwyidoi

Knowledge-Driven Network for Object Detection

Yundong Wu, Jiajia Liao, Yujun Liu, Kaiming Ding, Shimin Li, Zhilin Zhang, Guorong Cai, Jinhe Su
2021 Algorithms  
In this paper, we present a knowledge-driven network (KDNet)—a new architecture that can aggregate and model keypoint relations to augment object features for detection.  ...  In recent years, the concept of the object relationship model has become helpful for object detection and has been verified and realized in deep learning.  ...  A deficiency of a traditional anchor-free object detection framework. Shown in the first image, the anchor-free network detects two riders as the same object.  ... 
doi:10.3390/a14070195 fatcat:cfnhqungzbgkzkmebab2a6dyny

CenterNet3D:An Anchor free Object Detector for Autonomous Driving [article]

Guojun Wang, Bin Tian, Yunfeng Ai, Tong Xu, Long Chen, Dongpu Cao
2020 arXiv   pre-print
Based on the center point, we propose an anchor-free CenterNet3D Network that performs 3D object detection without anchors.  ...  Existing one-stage 3D object detection methods can achieve real-time performance, however, they are dominated by anchor-based detectors which are inefficient and require additional post-processing.  ...  Below, we will briefly review onestage and two-stage 3D object detection, and then we emphasize possibly related anchor-free 3D object detection. A.  ... 
arXiv:2007.07214v3 fatcat:7z3fxehs4rglbeymvftj5csg34

DAFNe: A One-Stage Anchor-Free Deep Model for Oriented Object Detection [article]

Steven Lang, Fabrizio Ventola, Kristian Kersting
2021 arXiv   pre-print
In this work, we present DAFNe: A Dense one-stage Anchor-Free deep Network for oriented object detection.  ...  The vast majority of the work in this area has focused on complex two-stage anchor-based approaches.  ...  Acknowledgements This work was supported by the Federal Ministry of Education and Research (BMBF) Competence Center for AI and Labour "kompAKI" (FKZ 02L19C150), the German Science Foundation (DFG, German  ... 
arXiv:2109.06148v2 fatcat:qve25gupwfhlppfhmfy7svnmvq

Axis Learning for Orientated Objects Detection in Aerial Images

Zhifeng Xiao, Linjun Qian, Weiping Shao, Xiaowei Tan, Kai Wang
2020 Remote Sensing  
In this article, we propose a new one-stage anchor-free method to detect orientated objects in per-pixel prediction fashion with less computational complexity.  ...  Most current methods for orientated object detection are anchor-based, which require considerable pre-defined anchors and are time consuming.  ...  In this paper, we propose a new one-stage anchor-free orientated objects detector for aerial images.  ... 
doi:10.3390/rs12060908 fatcat:ohce7ratdneqpgmvvjtzuozk6q

CentripetalNet: Pursuing High-quality Keypoint Pairs for Object Detection [article]

Zhiwei Dong, Guoxuan Li, Yue Liao, Fei Wang, Pengju Ren, Chen Qian
2020 arXiv   pre-print
In this paper, we propose CentripetalNet which uses centripetal shift to pair corner keypoints from the same instance.  ...  Furthermore, we explore instance segmentation on anchor-free detectors by equipping our CentripetalNet with a mask prediction module.  ...  Compared to the anchor-based approach, anchor-free detectors no longer need to preset anchor boxes. Mainly two types of anchor-free detectors are proposed.  ... 
arXiv:2003.09119v1 fatcat:wpl75jhj4fhqln4nlutojio4pm

Towards Accurate Scene Text Detection with Bidirectional Feature Pyramid Network

Dongping Cao, Jiachen Dang, Yong Zhong
2021 Symmetry  
In this paper, we propose a new Fully Convolutional One-Stage Object Detection (FCOS)-based text detection method that can robustly detect multioriented and multilingual text from natural scene images  ...  Our proposed text detector employs an anchor-free approach, unlike state-of-the-art text detectors that do not rely on a predefined anchor box.  ...  [32] proposed a new two-stage algorithm. In the first stage, the method predicts text instance locations by detecting and linking corners instead of traditional anchor points.  ... 
doi:10.3390/sym13030486 fatcat:haqr7qo4braw5obsyrhjgvyira

Detection of Multiscale Center Point Objects Based on Parallel Network

Hao Chen, Hong Zheng, Xiaolong Li
2020 Journal of Artificial Intelligence and Technology  
Though resembling the state-of-the-art two-stage detectors in overall object detection accuracy speed.  ...  On this basis, this paper proposes a multi-scale center point object detection method based on parallel network to further reduce the number of useless anchor boxes.  ...  Therefore, based on the two methods, a multiscale center point object detection method based on parallel network is proposed by this study.  ... 
doi:10.37965/jait.2020.0025 fatcat:vb2x5d54yjf3hplucfeosbo5ke

Object Detection in Aerial Images Using a Multiscale Keypoint Detection Network

Jinhe Su, Jiajiao Liao, Dujuan Gu, Zongyue Wang, Guorong Cai
2020 IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing  
In this article, a novel network, called the multiscale keypoint detection network (MKD-Net), is proposed to address these challenges.  ...  MKD-Net fuses multiscale layers to generate multiple feature maps for objects of different sizes. During the inference phase, both feature maps can be exploited for predicting corners.  ...  The feature selective anchor-free module for single-shot object detection [10] , guided anchoring [11] , the fully convolutional one-stage object detector [12] and FoveaBox [13] replace the bounding  ... 
doi:10.1109/jstars.2020.3044733 fatcat:smf7vg7bw5bczd4gxawsykahae

Joint 3D Proposal Generation and Object Detection from View Aggregation [article]

Jason Ku, Melissa Mozifian, Jungwook Lee, Ali Harakeh, Steven Waslander
2018 arXiv   pre-print
We present AVOD, an Aggregate View Object Detection network for autonomous driving scenarios.  ...  The proposed neural network architecture uses LIDAR point clouds and RGB images to generate features that are shared by two subnetworks: a region proposal network (RPN) and a second stage detector network  ...  Proposal Free Single Shot Detectors: Single shot object detectors have also been proposed as RPN free architectures for the 3D object detection task.  ... 
arXiv:1712.02294v4 fatcat:hdz4n3dkcncjlm63yui4asadku

MatrixNets: A New Scale and Aspect Ratio Aware Architecture for Object Detection [article]

Abdullah Rashwan, Rishav Agarwal, Agastya Kalra, Pascal Poupart
2020 arXiv   pre-print
First, we apply xNets on anchor-based object detection, for which we predict object centers and regress the top-left and bottom-right corners.  ...  Second, we use MatrixNets for corner-based object detection by predicting top-left and bottom-right corners. Each corner predicts the center location of the object.  ...  Two-stage detectors leverage a region proposal network to find a fixed number of object candidates. Then a second network is used to predict a score for each candidate and to refine its bounding box.  ... 
arXiv:2001.03194v1 fatcat:s4qnhhg4qbgyvjahwm5cp6ae6m

Scope Head for Accurate Localization in Object Detection [article]

Geng Zhan, Dan Xu, Guo Lu, Wei Wu, Chunhua Shen, Wanli Ouyang
2020 arXiv   pre-print
Existing anchor-based and anchor-free object detectors in multi-stage or one-stage pipelines have achieved very promising detection performance.  ...  Besides, an inherit anchor selection score is learned to indicate the localization quality of the detection result, and we propose to better represent the confidence of a detection box by combining the  ...  For two-stage detectors [23, 5, 6] , the direct regression is usually performed in region proposal generation and final object localization.  ... 
arXiv:2005.04854v2 fatcat:zfd5kcvlkfefnp27fb2uznavam

Anchor-free object detection with mask attention

He Yang, Beibei Fan, Lingling Guo
2020 EURASIP Journal on Image and Video Processing  
Another optimization strategy uses a similar two-stage object detection method.  ...  The anchor-free method based on key point detection has made great progress. However, the anchor-free method is too dependent on using a convolutional network to generate a rough heatmap.  ...  The authors express sincere appreciation to the anonymous referees for their helpful comments to improve the quality of the paper.  ... 
doi:10.1186/s13640-020-00517-3 fatcat:rmmpnd3iizgodnxszwvstnwpf4

Translational Symmetry-Aware Facade Parsing for 3D Building Reconstruction [article]

Hantang Liu, Wentong Li, Jianke Zhu
2021 arXiv   pre-print
In contrast to conventional semantic segmentation or bounding box prediction, we propose a novel scheme to fuse segmentation with anchor-free detection in a single stage network, which enables the efficient  ...  aided design, and city generation for digital entertainments.  ...  Fig. 2 : 2 Our proposed network architecture consists of two branches. One is for semantic segmentation, another one branch is for instance detection based anchor-free framework.  ... 
arXiv:2106.00912v1 fatcat:gdoubjs5ind45ktvphzs63mhoi

A Light-Weight Object Detection Framework with FPA Module for Optical Remote Sensing Imagery [article]

Xi Gu, Lingbin Kong, Zhicheng Wang, Jie Li, Zhaohui Yu, Gang Wei
2020 arXiv   pre-print
In this paper, we propose an efficient anchor free object detector, CenterFPANet. To pursue speed, we use a lightweight backbone and introduce the asymmetric revolution block.  ...  With the development of remote sensing technology, the acquisition of remote sensing images is easier and easier, which provides sufficient data resources for the task of detecting remote sensing objects  ...  The anchor free regression algorithm uses a single-stage object detection model for prediction, simplifying the two steps of object detection.  ... 
arXiv:2009.03063v1 fatcat:c7skjv4d7rbqndv3p4rle3b2vq
« Previous Showing results 1 — 15 out of 9,406 results