检索结果-内蒙古大学图书馆

Adaptive leader-follower formation control and obstacle avoidance via deep reinforcement learning

学校读者我要写书评

暂无评论

arXiv 2019年

作者： Zhou, Yanlin Lu, Fan Pu, George Ma, Xiyao Sun, Runhan Chen, Hsi-Yuan Li, Xiaolin Wu, Dapeng National Science Foundation Center for Big Learning Large-scale Intelligent Systems Laboratory Department of Electrical and Computer Engineering University of Florida GainesvilleFL United States Department of Mechanical and Aerospace Engineering University of Florida GainesvilleFL United States Amazon Robotics North Reading AI Institute Tongdun Technology

We propose a deep reinforcement learning (DRL) methodology for the tracking, obstacle avoidance, and formation control of nonholonomic robots. By separating vision-based control into a perception module and a controller module, we can train a DRL agent without sophisticated physics or 3D modeling. In addition, the modular framework averts daunting retrains of an image-to-action end-to-end neural network, and provides flexibility in transferring the controller to different robots. First, we train a convolutional neural network (CNN) to accurately localize in an indoor setting with dynamic foreground/background. Then, we design a new DRL algorithm named Momentum Policy Gradient (MPG) for continuous control tasks and prove its convergence. We also show that MPG is robust at tracking varying leader movements and can naturally be extended to problems of formation control. Leveraging reward shaping, features such as collision and obstacle avoidance can be easily integrated into a DRL controller. Copyright © 2019, The Authors. All rights reserved.

关键词： Deep learning

SCRDet: Towards More Robust Detection for Small, Cluttered and Rotated Objects

学校读者我要写书评

暂无评论

SCRDet: Towards More Robust Detection for Small, Cluttered a...

International Conference on computer Vision (ICCV)

作者： Xue Yang Jirui Yang Junchi Yan Yue Zhang Tengfei Zhang Zhi Guo Xian Sun Kun Fu NIST Institute of Electronics Beijing (Suzhou) China University of Chinese Academy of Sciences Beijing China Department of Computer Science and Engineering Shanghai Jiao Tong University MoE Key Lab of Artificial Intelligence AI Institute Shanghai Jiao Tong University

ISBN: (数字)9781728148038

ISBN: (纸本)9781728148045

Object detection has been a building block in computer vision. Though considerable progress has been made, there still exist challenges for objects with small size, arbitrary direction, and dense distribution. Apart from natural images, such issues are especially pronounced for aerial images of great importance. This paper presents a novel multi-category rotation detector for small, cluttered and rotated objects, namely SCRDet. Specifically, a sampling fusion network is devised which fuses multi-layer feature with effective anchor sampling, to improve the sensitivity to small objects. Meanwhile, the supervised pixel attention network and the channel attention network are jointly explored for small and cluttered object detection by suppressing the noise and highlighting the objects feature. For more accurate rotation estimation, the IoU constant factor is added to the smooth L1 loss to address the boundary problem for the rotating bounding box. Extensive experiments on two remote sensing public datasets DOTA, NWPU VHR-10 as well as natural image datasets COCO, VOC2007 and scene text data ICDAR2015 show the state-of-the-art performance of our detector. The code and models will be available at https://***/DetectionTeamUCAS.

关键词： Detectors Feature extraction Object detection Remote sensing Training Convolution Semantics

NTIRE 2021 Challenge on Image Deblurring

学校读者我要写书评

暂无评论

arXiv 2021年

作者： Nah, Seungjun Son, Sanghyun Lee, Suyoung Timofte, Radu Lee, Kyoung Mu Chen, Liangyu Zhang, Jie Lu, Xin Chu, Xiaojie Chen, Chengpeng Xiong, Zhiwei Xu, Ruikang Xiao, Zeyu Huang, Jie Zhang, Yueyi Xi, Si Wei, Jia Bai, Haoran Cheng, Songsheng Wei, Hao Sun, Long Tang, Jinhui Pan, Jinshan Lee, Donghyeon Lee, Chulhee Kim, Taesung Wang, Xiaobing Zhang, Dafeng Pan, Zhihong Lin, Tianwei Wu, Wenhao He, Dongliang Li, Baopu Li, Boyun Xi, Teng Zhang, Gang Liu, Jingtuo Han, Junyu Ding, Errui Tao, Guangpin Chu, Wenqing Cao, Yun Luo, Donghao Tai, Ying Lu, Tong Wang, Chengjie Li, Jilin Huang, Feiyue Chen, Hanting Chen, Shuaijun Guo, Tianyu Wang, Yunhe Zamir, Syed Waqas Arora, Aditya Khan, Salman Hayat, Munawar Khan, Fahad Shahbaz Shao, Ling Zuo, Yushen Ou, Yimin Chai, Yuanjun Shi, Lei Liu, Shuai Lei, Lei Feng, Chaoyu Zeng, Kai Yao, Yuying Liu, Xinran Zhang, Zhizhou Huang, Huacheng Zhang, Yunchen Jiang, Mingchao Zou, Wenbin Miao, Si Kim, Yangwoo Sun, Yuejin Deng, Senyou Ren, Wenqi Cao, Xiaochun Wang, Tao Suin, Maitreya Rajagopalan, A.N. Duong, Vinh Van Nguyen, Thuc Huu Yim, Jonghoon Jeon, Byeungwoo Li, Ru Xie, Junwei Han, Jong-Wook Choi, Jun-Ho Kim, Jun-Hyuk Lee, Jong-Seok Zhang, Jiaxin Peng, Fan Svitov, David Pakulich, Dmitry Kim, Jaeyeob Jeong, Jechang Department of ECE ASRI SNU Korea Republic of Computer Vision Lab ETH Zurich Switzerland University of Science and Technology of China China Megvii China Fudan University China Peking University China Netease Games AI Lab Nanjing University of Science and Technology China Guilin University of Electronic Technology China Samsung Electronics Co. Ltd Sunmoon University Asan Korea Republic of Samsung Research China Beijing China Baidu Research United States Department of Computer Vision Technology Baidu Inc China Fudan University China Megvii China Nanjing University China Tencent Noah's Ark Lab Huawei Technologies Co. Ltd Inception Institute of Artificial Intelligence Tsinghua University Beijing China North China University of Technology China Xiaomi South China University of Technology China Lab of Image Science and Technology Southeast University China China Design Group Co. Ltd China JOYY AI GROUP Fujian Normal University China Shanghai Advanced Research Institute Chinese Academy of Sciences China Institute of Information Engineering Chinese Academy of Sciences China Huawei Noah's Ark Lab Indian Institute of Technology Madras India Department of ECE Sungkyunkwan University Korea Republic of Fuzhou University China Imperial Vision Co. Ltd School of Integrated Technology Yonsei University Korea Republic of Expasoft LLC Institute of Automation and Electrometry The SB RAS Image Communication & Signal Processing Laboratory Hanyang University Korea Republic of

Motion blur is a common photography artifact in dynamic environments that typically comes jointly with the other types of degradation. This paper reviews the NTIRE 2021 Challenge on Image Deblurring. In this challenge report, we describe the challenge specifics and the evaluation results from the 2 competition tracks with the proposed solutions. While both the tracks aim to recover a high-quality clean image from a blurry image, different artifacts are jointly involved. In track 1, the blurry images are in a low resolution while track 2 images are compressed in JPEG format. In each competition, there were 338 and 238 registered participants and in the final testing phase, 18 and 17 teams competed. The winning methods demonstrate the state-of-the-art performance on the image deblurring task with the jointly combined artifacts. © 2021, CC BY-NC-ND.

关键词： Image enhancement

Revisiting metric learning for few-shot image classification

学校读者我要写书评

暂无评论

arXiv 2019年

作者： Li, Xiaomeng Yu, Lequan Fu, Chi-Wing Fang, Meng Heng, Pheng-Ann Department of Computer Science and Engineering Chinese University of Hong Kong Hong Kong Shenzhen Key Laboratory of Virtual Reality and Human Interaction Technology Shenzhen Institutes of Advanced Technology Chinese Academy of Sciences China Tencent AI Lab United States

The goal of few-shot learning is to recognize new visual concepts with just a few amount of labeled samples in each class. Recent effective metric-based few-shot approaches employ neural networks to learn a feature similarity comparison between query and support examples. However, the importance of feature embedding, i.e., exploring the relationship among training samples, is neglected. In this work, we present a simple yet powerful baseline for few-shot classification by emphasizing the importance of feature embedding. Specifically, we revisit the classical triplet network from deep metric learning, and extend it into a deep K-tuplet network for few-shot learning, utilizing the relationship among the input samples to learn a general representation learning via episode-training. Once trained, our network is able to extract discriminative features for unseen novel categories and can be seamlessly incorporated with a non-linear distance metric function to facilitate the few-shot classification. Our result on the miniImageNet benchmark outperforms other metric-based few-shot classification methods. More importantly, when evaluated on completely different datasets (Caltech-101, CUB-200, Stanford Dogs and Cars) using the model trained with miniImageNet, our method significantly outperforms prior methods, demonstrating its superior capability to generalize to unseen classes. Copyright © 2019, The Authors. All rights reserved.

关键词： Image classification

Probing contextualized sentence representations with visual awareness

学校读者我要写书评

暂无评论

arXiv 2019年

作者： Zhang, Zhuosheng Wang, Rui Chen, Kehai Utiyama, Masao Sumita, Eiichiro Zhao, Hai Department of Computer Science and Engineering Shanghai Jiao Tong University Key Laboratory of Shanghai Education Commission for Intelligent Interaction and Cognitive Engineering Shanghai Jiao Tong University Shanghai China MoE Key Lab of Artificial Intelligence AI Institute Shanghai Jiao Tong University Shanghai China

We present a universal framework to model contextualized sentence representations with visual awareness that is motivated to overcome the shortcomings of the multimodal parallel data with manual annotations. For each sentence, we first retrieve a diversity of images from a shared cross-modal embedding space, which is pre-trained on a large-scale of text-image pairs. Then, the texts and images are respectively encoded by transformer encoder and convolutional neural network. The two sequences of representations are further fused by a simple and effective attention layer. The architecture can be easily applied to text-only natural language processing tasks without manually annotating multimodal parallel corpora. We apply the proposed method on three tasks, including neural machine translation, natural language inference and sequence labeling and experimental results verify the effectiveness. Copyright © 2019, The Authors. All rights reserved.

关键词： Natural language processing systems

Provably efficient maximum entropy exploration

学校读者我要写书评

暂无评论

arXiv 2018年

作者： Hazan, Elad Kakade, Sham M. Singh, Karan van Soest, Abby Google AI Princeton Department of Computer Science Princeton University Allen School of Computer Science and Engineering University of Washington Department of Statistics University of Washington

Suppose an agent is in a (possibly unknown) Markov Decision Process in the absence of a reward signal, what might we hope that an agent can efficiently learn to do? This work studies a broad class of objectives that are defined solely as functions of the state-visitation frequencies that are induced by how the agent behaves. For example, one natural, intrinsically defined, objective problem is for the agent to learn a policy which induces a distribution over state space that is as uniform as possible, which can be measured in an entropic sense. We provide an efficient algorithm to optimize such such intrinsically defined objectives, when given access to a black box planning oracle (which is robust to function approximation). Furthermore, when restricted to the tabular setting where we have sample based access to the MDP, our proposed algorithm is provably efficient, both in terms of its sample and computational complexities. Key to our algorithmic methodology is utilizing the conditional gradient method (a.k.a. the Frank-Wolfe algorithm) which utilizes an approximate MDP solver. Copyright © 2018, The Authors. All rights reserved.

关键词： Markov processes

Complementary time-frequency domain networks for dynamic parallel MR image reconstruction

学校读者我要写书评

暂无评论

arXiv 2020年

作者： Qin, Chen Duan, Jinming Hammernik, Kerstin Schlemper, Jo Küstner, Thomas Botnar, René Prieto, Claudia Price, Anthony N. Hajnal, Joseph V. Rueckert, Daniel Institute for Digital Communications School of Engineering University of Edinburgh Edinburgh United Kingdom Department of Computing Imperial College London London United Kingdom School of Computer Science University of Birmingham Birmingham United Kingdom Institute for AI and Informatics Klinikum Rechts der Isar Technical University of Munich Munich Germany Hyperfine Research Inc. GuilfordCT United States School of Biomedical Engineering and Imaging Sciences King's College London St. Thomas' Hospital London United Kingdom Department of Diagnostic and Interventional Radiology Medical Image and Data Analysis University Hospital of Tuebingen Tuebingen Germany

Purpose To introduce a novel deep learning based approach for fast and high-quality dynamic multi-coil MR reconstruction by learning a complementary time-frequency domain network that exploits spatio-temporal correlations simultaneously from complementary domains. Theory and Methods Dynamic parallel MR image reconstruction is formulated as a multi-variable minimisation problem, where the data is regularised in combined temporal Fourier and spatial (x-f ) domain as well as in spatio-temporal image (x-t) domain. An iterative algorithm based on variable splitting technique is derived, which alternates among signal de-aliasing steps in x-f and x-t spaces, a closed-form point-wise data consistency step and a weighted coupling step. The iterative model is embedded into a deep recurrent neural network which learns to recover the image via exploiting spatio-temporal redundancies in complementary domains. Results Experiments were performed on two datasets of highly undersampled multi-coil short-axis cardiac cine MRI scans. Results demonstrate that our proposed method outperforms the current state-of-the-art approaches both quantitatively and qualitatively. The proposed model can also generalise well to data acquired from a different scanner and data with pathologies that were not seen in the training set. Conclusion The work shows the benefit of reconstructing dynamic parallel MRI in complementary time-frequency domains with deep neural networks. The method can effectively and robustly reconstruct high-quality images from highly undersampled dynamic multi-coil data (16× and 24× yielding 15s and 10s scan times respectively) with fast reconstruction speed (2.8s). This could potentially facilitate achieving fast single-breath-hold clinical 2D cardiac cine imaging. © 2020, CC BY.

关键词： Magnetic resonance imaging

Learning to multitask

学校读者我要写书评

暂无评论

arXiv 2018年

作者： Zhang, Yu Wei, Ying Yang, Qiang Department of Computer Science and Engineering Hong Kong University of Science and Technology Ai Lab Tencent United States

Multitask learning has shown promising performance in many applications and many multitask models have been proposed. In order to identify an effective multitask model for a given multitask problem, we propose a learning framework called learning to multitask (L2MT). To achieve the goal, L2MT exploits historical multitask experience which is organized as a training set consists of several tuples, each of which contains a multitask problem with multiple tasks, a multitask model, and the relative test error. Based on such training set, L2MT first uses a proposed layerwise graph neural network to learn task embeddings for all the tasks in a multitask problem and then learns an estimation function to estimate the relative test error based on task embeddings and the representation of the multitask model based on a unified formulation. Given a new multitask problem, the estimation function is used to identify a suitable multitask model. Experiments on benchmark datasets show the effectiveness of the proposed L2MT framework. Copyright © 2018, The Authors. All rights reserved.

关键词： Embeddings

Hierarchical contextualized representation for named entity recognition

学校读者我要写书评

暂无评论

arXiv 2019年

作者： Luo, Ying Xiao, Fengshun Zhao, Hai Department of Computer Science and Engineering Shanghai Jiao Tong University Key Laboratory of Shanghai Education Commission for Intelligent Interaction and Cognitive Engineering Shanghai Jiao Tong University Shanghai China MoE Key Lab of Artificial Intelligence AI Institute Shanghai Jiao Tong University Shanghai China

Named entity recognition (NER) models are typically based on the architecture of Bi-directional LSTM (BiLSTM). The constraints of sequential nature and the modeling of single input prevent the full utilization of global information from larger scope, not only in the entire sentence, but also in the entire document (dataset). In this paper, we address these two deficiencies and propose a model augmented with hierarchical contextualized representation: Sentence-level representation and document-level representation. In sentencelevel, we take different contributions of words in a single sentence into consideration to enhance the sentence representation learned from an independent BiLSTM via label embedding attention mechanism. In document-level, the key-value memory network is adopted to record the document-aware information for each unique word which is sensitive to similarity of context information. Our two-level hierarchical contextualized representations are fused with each input token embedding and corresponding hidden state of BiLSTM, respectively. The experimental results on three benchmark NER datasets (CoNLL-2003 and Ontonotes 5.0 English datasets, CoNLL-2002 Spanish dataset) show that we establish new state-of-the-art results. Copyright © 2019, The Authors. All rights reserved.

关键词： Embeddings