16 Hits in 5.4 sec

FMT:Fusing Multi-task Convolutional Neural Network for Person Search [article]

Sulan Zhai, Shunqiang Liu, Xiao Wang, Jin Tang
2020 arXiv   pre-print
In this paper, we propose a fusing multi-task convolutional neural network(FMT-CNN) to tackle the correlation and heterogeneity of detection and re-identification with a single convolutional neural network  ...  We employ person labels in region proposal network to produce features for person re-identification and person detection network, which can improve the accuracy of detection and re-identification simultaneously  ...  Proposed Approach In this section, we present the details of our proposed fusing multi-task convolutional neural network for person search(FMT-CNN).  ... 
arXiv:2003.00406v1 fatcat:ssbcvqpwbfc4zeyjgdinfzwape

Person search: New paradigm of person re-identification: A survey and outlook of recent works

Khawar Islam
2020 Image and Vision Computing  
Specially, convolutional neural network (CNN) achieves breakthrough performance and extracts useful patterns and characteristics.  ...  In last few years, deep learning has played unremarkable role for the solution of re-identification problem. Deep learning shows incredible performance in person (re-ID) and search.  ...  PFFN [60] Proposed PFFN, the goal of network is to fuse multi-level convolutional neural network feature maps in topmost style.  ... 
doi:10.1016/j.imavis.2020.103970 fatcat:g2zuqww7tbdszkxrc2wkrfno2y

2021 Index IEEE Transactions on Vehicular Technology Vol. 70

2021 IEEE Transactions on Vehicular Technology  
The Author Index contains the primary entry for each item, listed under the first author's name.  ...  Masuho, J., +, TVT Sept. 2021 8468-8477 A Personalized Search Query Generating Method for Safety-Enhanced Vehicle-to-People Networks.  ...  ., +, TVT Sept. 2021 9244-9257 Information retrieval A Personalized Search Query Generating Method for Safety-Enhanced Vehi-cle-to-People Networks.  ... 
doi:10.1109/tvt.2022.3151213 fatcat:vzuzqu54irebpibzp3ykgy5nca

Perception, Planning, Control, and Coordination for Autonomous Vehicles

Scott Pendleton, Hans Andersen, Xinxin Du, Xiaotong Shen, Malika Meghjani, You Eng, Daniela Rus, Marcelo Ang
2017 Machines  
Autonomous vehicles are expected to play a key role in the future of urban transportation systems, as they offer potential for additional safety, increased productivity, greater accessibility, better road  ...  Different from these two approaches, Mohan [121] proposed a novel architecture that integrated the CNN with deep de-convolutional neural networks.  ...  In [65] , VoxNet was proposed, which implemented a 3D convolutional neural network to classify the 3D point cloud (in occupancy grid/volumetric representation).  ... 
doi:10.3390/machines5010006 fatcat:24h5nl5quve6fpt2wtdyf4qafe

Applications of Deep Neural Networks with Keras [article]

Jeff Heaton
2022 arXiv   pre-print
Deep learning is a group of exciting new technologies for neural networks.  ...  This course will introduce the student to classic neural network structures, Convolution Neural Networks (CNN), Long Short-Term Memory (LSTM), Gated Recurrent Neural Networks (GRU), General Adversarial  ...  Useful for tasks such as connecting RNNs and convolutional networks. • RepeatVector -Repeats the input n times.  ... 
arXiv:2009.05673v5 fatcat:h3jghqylwrbfvfglmwutlfpmay

Real-time Autonomous Robot Navigation System with Collision Avoidance for NAO Robot

Oluwadamilola I Saka
It integrates image and depth-based environmental perception, path planning, and walking control of the NAO humanoid robot for navigation in unknown terrains.  ...  The path to move from the current position to the destination is solved using a search algorithm. The robot moves along the defined path by the walking controller module.  ...  The OverFeat, released in 2013, used classic convolutional neural network (CNN) for the task of image feature extraction and classification [42] .  ... 
doi:10.25417/uic.17026154 fatcat:wdcujkf7urfntjzylnk35qewj4

Introduction [chapter]

2016 Music Data Analysis  
Convolutional neural networks (CNNs) are a particularly successful class of deep networks.  ...  Examples of such models are Support Vector Machines [5] or Convolutional Neural Networks [9] (see also Chapter 12).  ...  Later, [19] described a method to extract four moods (anger, fear, hap piness, sadness) from tempo and articulation features using a neural network.  ... 
doi:10.1201/9781315370996-5 fatcat:avooqogcpnbjngqmzuonil3exq

Development of a real-time classifier for the identification of the Sit-To-Stand motion pattern

The optimised architecture combined convolutional and recurrent neural networks into a hybrid approach and was able to correctly identify the four STS phases, both under SP and CT movements, relying on  ...  The overall accuracy estimate (median [95% confidence intervals]) for the hybrid architecture was 96.09 [95.37 - 96.56] in SP trials and 95.74 [95.39 ? 96.21] in CT trials.  ...  Convolutional neural networks As we highlighted in the previous section, RNNs are commonly considered the starting point (if not the favourite solution) for sequential predictive tasks [207] .  ... 
doi:10.15167/job-mirko_phd2021-07-15 fatcat:qlhdpcdvfzg43k2tj55sbg5nsi

Monitoring reperfused myocardial infarction with delayed left ventricular systolic dysfunction in rabbits by longitudinal imaging

Yuanbo Feng, Bianca Hemmeryckx, Liesbeth Frederix, Marleen Lox, Jun Wu, Ward Heggermont, Hua Rong Lu, David Gallacher, Raymond Oyen, H Roger Lijnen, Yicheng Ni
2018 Quantitative Imaging in Medicine and Surgery  
An experimental imaging platform for longitudinal monitoring and evaluation of cardiac morphology-function changes has been long desired.  ...  Seven weeks post MI, animals were sacrificed and heart tissues were processed for histopathological staining. The success rate of surgical operation was 87.27%.  ...  expert technical assistance, and Jie Yu for making histological preparations.  ... 
doi:10.21037/qims.2018.09.05 pmid:30306056 pmcid:PMC6177364 fatcat:cqroeb5ch5enlh76obrg4rlyea

Efficient in-hardware compression of on-chip data

Amin Ghasemazar
We first pro- pose Channeleon, which tackles the problem of compressing the activation maps in deep neural networks (DNNs) at inference time.  ...  SparTen: A Sparse Tensor Accelerator for Convolutional Neural Networks. In MICRO, 2019.  ...  Shufflenet: An extremely efficient convolutional neural network for mobile devices.  ... 
doi:10.14288/1.0404515 fatcat:nxtj5xz4yffm7inyjyp7k6caem

Full Issue PDF

2020 JACC Cardiovascular Imaging  
Patients were re-evaluated 6 weeks after discharge and prospectively followed up for the composite endpoint of heart failure readmission and all-cause mortality.  ...  Lower peak atrial longitudinal strain values after decongestion were associated with increased risk for the composite endpoint of heart failure and mortality (p < 0.019).  ...  The authors are solely responsible for the design and conduct of this study; all study analyses; and the drafting and editing of the paper and its final contents.  ... 
doi:10.1016/s1936-878x(20)30301-6 fatcat:w765zpyw4neohdsuytjzyy4gtm

2020 Western Medical Research Conference

2019 Journal of Investigative Medicine  
At each of the K-9 folds, 88% of data was used to train the neural network (including 216 control and 513 CAH images) and the rest was used to evaluate the trained neural network model (including 70 CAH  ...  Here, we examine this question through quantifying the neural network (NN) performance of prostate organ segmentation against the number of samples.  ...  His labs were notable for an elevated C-reactive protein and mildly elevated ESR but otherwise normal CBC and CMP. Pediatric Surgery was consulted due to concern for appendicitis.  ... 
doi:10.1136/jim-2019-wmrc fatcat:cw2dtzk5vzc2bmd4wbujao6igm

33rd Annual Meeting & Pre-Conference Programs of the Society for Immunotherapy of Cancer (SITC 2018)

2018 Journal for ImmunoTherapy of Cancer  
O8 The GAPVAC approach of actively personalized peptide vaccination for patients with newly diagnosed glioblastoma Norbert Hilf, PhD 1 Acknowledgements We thank the following people for their technical  ...  assistance:In vivo team for animal studies, Protein sciences team for protein production Ethics Approval Murine studies were conducted under a U.K.  ...  Results We first present AI-MHC, an applied deep convolutional neural network for class-specific MHC binding algorithm that achieves state-ofthe-art performance in both Class and Class II predictions.  ... 
doi:10.1186/s40425-018-0423-x pmid:30400822 pmcid:PMC6220479 fatcat:vnbdtqcvqrfdtlsgirjycjot6e

Abstractband der DGTI 2022

2022 Transfusion Medicine and Hemotherapy  
Conclusion: Der Vortrag gibt einen Überblick über die Herstellung von FMT-Präparaten und die eingeführte Eurocode-Systematik.  ...  Der fäkale Mikrobiota-Transfer (FMT) ist eine neue therapeutische Option, insbesondere für die Behandlung von rezidivierenden Clostridioides difficile Infektionen (rCDI). Unter der Leitung von Prof.  ...  We have combined statistical power of flow morphometry using an in-house developed microfluidic system with high image classification power of a convolutional neural network (CNN) to assess changes in  ... 
doi:10.1159/000525886 fatcat:uaskpccmondl7kfarjmktxifi4

Music Encoding Conference Proceedings 2021. 19–22 July, 2021 University of Alicante (Spain): Onsite & Online. Edited by Stefan Münnich and David Rizo [article]

HC User, Stefan Münnich, David Rizo
The MELD framework was created, with David Weigl as lead developer, as part of the Fusing Audio and Semantic Technologies for Intelligent Music Production and Acknowledgments The author likes to express  ...  Bohl for his contributions to MPM's conception, Peter Stadler for his ODD support and Simon Waloschek for developing the doc_generator.  ...  These approaches have in common that they make heavy use of deep convolutional neural networks.  ... 
doi:10.17613/fc1c-mx52 fatcat:yq2zdtkf2bbqpoyls2spog6gky
« Previous Showing results 1 — 15 out of 16 results