162 Hits in 11.8 sec

A Classification Supervised Auto-Encoder Based on Predefined Evenly-Distributed Class Centroids [article]

Qiuyu Zhu, Ruixin Zhang
2020 arXiv   pre-print
In this paper, a new autoencoder model - classification supervised autoencoder (CSAE) based on predefined evenly-distributed class centroids (PEDCC) is proposed.  ...  Based on the basic structure of the universal autoencoder, we realized the comprehensive optimal results of encoding, decoding, classification, and good model generalization performance at the same time  ...  As far as we know, prior to this article, there was no method of using predefined class centroids to A Classification Supervised Auto-Encoder Based on Predefined Evenly-Distributed Class Centroids  ... 
arXiv:1902.00220v3 fatcat:6q26owjk4ven5nskicjisciqwi

An Image Clustering Auto-Encoder Based on Predefined Evenly-Distributed Class Centroids and MMD Distance [article]

Qiuyu Zhu, Zhengyong Wang
2019 arXiv   pre-print
The algorithm uses PEDCC (Predefined Evenly-Distributed Class Centroids) as the clustering centers of the images, which ensures the inter-class distance of latent features is maximal, and adds data distribution  ...  In addition, we can use the pre-defined PEDCC class centers, and the decoding module of the auto-encoder to clearly generate the samples of each class. The code can be downloaded at xxx!  ...  [46] proposed the Classification Supervised Auto-Encoder (CSAE), which used the predefined uniform distribution class centers to realize the classification function of the autoencoder, and generated  ... 
arXiv:1906.03905v1 fatcat:wbjaha3gtzdzxefgua2nvhb5hu

Dual Supervised Autoencoder Based Trajectory Classification Using Enhanced Spatio-Temporal Information

Sichong Lu, Ying Xia
2020 IEEE Access  
based on the Predefined Evenly-Distributed Class Centroids (PEDCC) algorithm [26] . 3) A neural network consisting of two autoencoders is designed, one of which learns the high-level features of RPs  ...  DUAL CNN BASED SUPERVISED AUTOENCODER USING PREDEFINED CLASS CENTROIDS From the last section, we can conclude that AF and RP not only enhance spatio-temporal information but also have a better data structure  ... 
doi:10.1109/access.2020.3026110 fatcat:rckw43yesjfpvfizj25clkummu

Self-supervised Cross-silo Federated Neural Architecture Search [article]

Xinle Liang, Yang Liu, Jiahuan Luo, Yuanqin He, Tianjian Chen, Qiang Yang
2021 arXiv   pre-print
Then, parties collaboratively improve the local optimal architecture in a VFL framework with supervision.  ...  In the proposed framework, each party first conducts NAS using self-supervised approach to find a local optimal architecture with its own data.  ...  Firstly, we generate multi-view images based on the approaches in [66] . Then we distribute the images evenly to different parties.  ... 
arXiv:2101.11896v2 fatcat:52fszbiqj5eepenxjlifvr37ge

Interactive online learning for obstacle classification on a mobile robot

Viktor Losing, Barbara Hammer, Heiko Wersing
2015 2015 International Joint Conference on Neural Networks (IJCNN)  
We employ a cost-function-based learning vector quantization approach and introduce a new insertion strategy optimizing a cost-function based on a subset of samples.  ...  The model is based on learning vector quantization, approaching the stability-plasticity problem of incremental learning by adaptive insertions of representative vectors.  ...  Whether a sample is promoted to a prototype or not is solely based on its distance to nodes of other classes.  ... 
doi:10.1109/ijcnn.2015.7280610 dblp:conf/ijcnn/LosingHW15 fatcat:jrex67hekvbpxn7qzj5vqgat7q

Contradistinguisher: A Vapnik's Imperative to Unsupervised Domain Adaptation [article]

Sourabh Balgi, Ambedkar Dukkipati
2021 arXiv   pre-print
in a supervised way on the source domain.  ...  Recent domain adaptation works rely on an indirect way of first aligning the source and target domain distributions and then train a classifier on the labeled source domain to classify the target domain  ...  based on maximum entropy distribution to fit the observed data by identifying the patterns in the observed data.  ... 
arXiv:2005.14007v3 fatcat:nkofjc3pbjdt5aq7arhdh3bmj4

Multimodal Transformer for Automatic 3D Annotation and Object Detection [article]

Chang Liu, Xiaoyan Qian, Binxiao Huang, Xiaojuan Qi, Edmund Lam, Siew-Chong Tan, Ngai Wong
2022 arXiv   pre-print
To alleviate the pervasive sparsity problem that hinders existing autolabelers, MTrans densifies the sparse point clouds by generating new 3D points based on 2D image information.  ...  MTrans can also be extended to improve the accuracy for 3D object detection, resulting in a remarkable 89.45% AP on KITTI hard samples. Codes are at .  ...  SDF [32] uses predefined CAD models to estimate the 3D geometry of cars detected in 2D images. VS3D [17] generates 3D proposals based on cloud density with an unsupervised UPM module.  ... 
arXiv:2207.09805v1 fatcat:j4fdx5w7e5dpfa6pyj2zndynn4

Mutation testing: Clustering the mutants

Basarat, Oprescu, Biesaart, Ana Oprescu
2021 Zenodo  
Based on this research we selected a mutation testing tool to use in our research. We identified characteristics to represent the mutants such that we can cluster them.  ...  One of these techniques is the clustering of mutants. By clustering mutants we can execute less mutants to reduce the cost. Our research consists of two parts in which we cluster mutants.  ...  Ana Oprescu, for her guidance ever since my first semester on the University of Amsterdam. Thank you Ana, for being my supervisor throughout this thesis and being so involved with your students.  ... 
doi:10.5281/zenodo.5750688 fatcat:ypkglhn35nec7fi7ht2ijxipnm

Learning image context for segmentation of the prostate in CT-guided radiotherapy

Wei Li, Shu Liao, Qianjin Feng, Wufan Chen, Dinggang Shen
2012 Physics in Medicine and Biology  
In this paper, an online-learning and patient-specific classification method based on the location-adaptive image context is presented to deal with all these challenging issues and achieve the precise  ...  The proposed learning-based prostate segmentation method has been extensively evaluated on 161 images of 11 patients, each with more than nine daily treatment three-dimensional CT images.  ...  To better approximate the marginal distribution, an auto-context model is proposed, in which context features are extracted from the classification probability map F t produced by a traditional classifier  ... 
doi:10.1088/0031-9155/57/5/1283 pmid:22343071 pmcid:PMC3378724 fatcat:n7vgyk5qlbchbmt7zzezp3s6wu

LiveCellMiner : A new tool to analyze mitotic progression

Daniel Moreno-Andrés, Anuk Bhattacharyya, Anja Scheufen, Johannes Stegmaier
2022 PLOS ONE 17(7)  
The software tool allows analysis across high-content data sets obtained in different platforms, in a quantitative and unbiased manner.  ...  The general need for dedicated experiment-specific user-annotated training sets and experiment-specific user-defined segmentation parameters remains a major bottleneck for fully automating the analysis  ...  PLOS ONE  ... 
doi:10.18154/rwth-2022-07737 fatcat:ipki4vor65adzpgdfmvopnpfuq

Fisher Kernel Temporal Variation-based Relevance Feedback for video retrieval

Ionuţ Mironică, Bogdan Ionescu, Jasper Uijlings, Nicu Sebe
2016 Computer Vision and Image Understanding  
This paper proposes a novel framework for Relevance Feedback based on the Fisher Kernel (FK).  ...  Specifically, we train a Gaussian Mixture Model (GMM) on the top retrieval results (without supervision) and use this to create a FK representation, which is therefore specialized in modelling the most  ...  [71] introduces a new encoding technique that generates a video representation based on CNN descriptors.  ... 
doi:10.1016/j.cviu.2015.10.005 fatcat:ch7sxtrnnjdv5jwfxsx5mrou3i

Machine Learning for Exploring Spatial Affordance Patterns [article]

Boyana Buyuklieva
2020 arXiv   pre-print
achieve a strong cluster-to-class evaluation.  ...  The thesis also includes an evaluation of the layout case studies with unsupervised learners, which showed that use could not be immediately reverse-engineered based solemnly on the VGA information to  ...  The initial intent of the work was to create a layout generating tool based on an expert-system trained on real examples that could suggest function/usage spaces very much like an auto-predict.  ... 
arXiv:2005.08106v1 fatcat:rx3cquprbvbmta6vahrqrgxlgu

Automatic Fabric Defect Detection Using Cascaded Mixed Feature Pyramid with Guided Localization

Wu, Zhang, Fang
2020 Sensors  
Stacked feature pyramid networks are set up to aggregate cross-scale defect patterns on interpolating mixed depth-wise block in stage one.  ...  The experiments show that the end-to-end architecture improves the occluded defect performance of region-based object detectors as compared with the current detectors.  ...  proposed the Region Proposal Network (RPN) [8] to generate proposals in a supervised way based on sliding convolution filters.  ... 
doi:10.3390/s20030871 pmid:32041348 pmcid:PMC7039386 fatcat:bw6ncpas5vcubjgejal2sfnera

Deeply Exploiting Long-Term View Dependency for 3D Shape Recognition

Yong Xu, Chaoda Zheng, Ruotao Xu, Yuhui Quan
2019 IEEE Access  
view aggregation module based on the bi-directional Long Short-Term Memory network.  ...  Incorporating the aggregation module into a standard convolutional network architecture, we develop an effective method for 3D shape classification and retrieval.  ...  To anylase the shape distribution of 3D objects, [51] uses a VAE (variational auto-encoder) to reconstruct full 3D shapes from voxelized single-views.  ... 
doi:10.1109/access.2019.2934650 fatcat:vo7jdyq7qnbnvok6uwhblgzkvm

Introduction to information retrieval

2009 ChoiceReviews  
But the outcome depends on the details of the distributed system; at least one thread of work has reached the opposite conclusion Barbosa 1998, Badue et al. 2001).  ...  The scheme discussed in this chapter, currently believed to be the best published scheme (achieving as few as 3 bits per link for encoding), is described in a series of papers by Boldi and Vigna (2004b  ...  of learning is called supervised learning since a "supervisor" (the SUPERVISED LEARNING human who defines the classes and labels training documents for each class) functions as a teacher directing the  ... 
doi:10.5860/choice.46-2715 fatcat:ruwoe46pgzcupjygnwbnit4z3u
« Previous Showing results 1 — 15 out of 162 results