435,023 Hits in 4.5 sec

Deep One-Class Classification Using Intra-Class Splitting [article]

Patrick Schlachter, Yiwen Liao, Bin Yang
2019 arXiv   pre-print
The method is based on splitting given data from one class into two subsets. In one-class classification, only samples of one normal class are available for training.  ...  This paper introduces a generic method which enables to use conventional deep neural networks as end-to-end one-class classifiers.  ...  Accordingly, an obvious research direction is to use deep learning methods for one-class classification. Indeed, there exist only few deep learning approaches to one-class classification.  ... 
arXiv:1902.01194v3 fatcat:gwlitnv6hzfwjdfihuxjeifb7a

Explainable Deep One-Class Classification [article]

Philipp Liznerski, Lukas Ruff, Robert A. Vandermeulen, Billy Joe Franks, Marius Kloft, Klaus-Robert Müller
2021 arXiv   pre-print
Deep one-class classification variants for anomaly detection learn a mapping that concentrates nominal samples in feature space causing anomalies to be mapped away.  ...  Finally, using FCDD's explanations we demonstrate the vulnerability of deep one-class classification models to spurious image features such as image watermarks.  ...  Explaining Deep One-Class Classification We review one-class classification and fully convolutional architectures before presenting our method.  ... 
arXiv:2007.01760v3 fatcat:rrexnt4icnhahl2bz2z6i4ohkm

Learning Deep Features for One-Class Classification [article]

Pramuditha Perera, Vishal M. Patel
2019 arXiv   pre-print
We propose a deep learning-based solution for the problem of feature learning in one-class classification.  ...  Extensive experiments on publicly available anomaly detection, novelty detection and mobile active authentication datasets show that the proposed Deep One-Class (DOC) classification method achieves significant  ...  Details of the proposed deep one-class classification method are given in Section III.  ... 
arXiv:1801.05365v2 fatcat:pdx5hycy4jdizopsavf3ihjjm4

DOC3-Deep One Class Classification using Contradictions [article]

Sauptik Dhar, Bernardo Gonzalez Torres
2022 arXiv   pre-print
We formalize this notion for the widely adopted one class large-margin loss, and propose the Deep One Class Classification using Contradictions (DOC3) algorithm.  ...  This paper introduces the notion of learning from contradictions (a.k.a Universum learning) for deep one class classification problems.  ...  For deep learning models, we optimize (6) over all the model parameters and refer to it as Deep One Class Classification using Contradictions (DOC 3 ).  ... 
arXiv:2105.07636v2 fatcat:6bhutr4eqng7lf4oibctloqwui

DROCC: Deep Robust One-Class Classification [article]

Sachin Goyal, Aditi Raghunathan, Moksh Jain, Harsha Vardhan Simhadri, Prateek Jain
2020 arXiv   pre-print
In this work, we propose Deep Robust One-Class Classification (DROCC) that is both applicable to most standard domains without requiring any side-information and robust to representation collapse.  ...  Classical approaches for one-class problems such as one-class SVM and isolation forest require careful feature engineering when applied to structured domains like images.  ...  Deep Robust One Class Classification: We now present our approach to unsupervised anomaly detection that we call Deep Robust One Class Classification (DROCC).  ... 
arXiv:2002.12718v2 fatcat:3vxztrzyrvfz3k7v7cjr2fsyqa

Simple and Effective Prevention of Mode Collapse in Deep One-Class Classification [article]

Penny Chong, Lukas Ruff, Marius Kloft, Alexander Binder
2020 arXiv   pre-print
In this work, we consider two regularizers to prevent hypersphere collapse in deep SVDD. The first regularizer is based on injecting random noise via the standard cross-entropy loss.  ...  Our proposed regularized variants of deep SVDD show encouraging results and outperform a prominent state-of-the-art method on a setup where the anomalies have no apparent geometrical structure.  ...  Recently [13] proposed an unsupervised one-class classification approach-deep SVDD, which jointly learns the feature representation of the data and a data-enclosing hypersphere.  ... 
arXiv:2001.08873v3 fatcat:bzxna4rtubb7pcv7qk24clfbji

Learning and Evaluating Representations for Deep One-class Classification [article]

Kihyuk Sohn, Chun-Liang Li, Jinsung Yoon, Minho Jin, Tomas Pfister
2021 arXiv   pre-print
We present a two-stage framework for deep one-class classification.  ...  Finally, we present visual explanations, confirming that the decision-making process of deep one-class classifiers is intuitive to humans.  ...  Following the success of deep learning [15] , deep one-class classifications [16, 17, 18] , which extend the discriminative one-class classification using trainable deep neural networks, have shown promising  ... 
arXiv:2011.02578v2 fatcat:y2m2foa7i5ayrmabwxbnjjqrn4

Deep Nearest Class Mean Model for Incremental Odor Classification

2018 IEEE Transactions on Instrumentation and Measurement  
Motivated by this concern, this paper proposes a Deep Nearest Class Mean (DNCM) model based on the deep learning framework and nearest class mean method.  ...  The proposed model not only leverages deep neural network to extract deep features, but is also able to dynamically integrate new classes over time.  ...  Another method of this research line is one vs. one [41] framework which trains ( − 1) 2 ⁄ binary classifier for N-way multi-class classification problem.  ... 
doi:10.1109/tim.2018.2863438 fatcat:nljephiwz5aqpfjlployy2np7y

Deep One-Class Classification via Interpolated Gaussian Descriptor [article]

Yuanhong Chen and Yu Tian and Guansong Pang and Gustavo Carneiro
2022 arXiv   pre-print
One-class classification (OCC) aims to learn an effective data description to enclose all normal training samples and detect anomalies based on the deviation from the data description.  ...  To address this issue, we introduce the interpolated Gaussian descriptor (IGD) method, a novel OCC model that learns a one-class Gaussian anomaly classifier trained with adversarially interpolated training  ...  In Table S3 , we only show the results of our approach because we could not find the class-level results for other approaches. On the class-level results for CIFAR10, on Tab.  ... 
arXiv:2101.10043v5 fatcat:x2lvlxhs4bbvrm7dvmwfolnsui

nanoDoc: RNA modification detection using Nanopore raw reads with Deep One-Class Classification [article]

Hiroki Ueda
2020 bioRxiv   pre-print
Current signal deviations caused by PTMs are analyzed via Deep One-Class Classification with a convolutional neural network.  ...  We also demonstrated a tentative classification of PTMs using unsupervised clustering.  ...  Detection of DNA base modifications by deep 304 recurrent neural network on Oxford Nanopore sequencing data.  ... 
doi:10.1101/2020.09.13.295089 fatcat:fsjip4s3abbtdmvassffgbjega

On the generalization of bayesian deep nets for multi-class classification [article]

Yossi Adi, Yaniv Nemcovsky, Alex Schwing, Tamir Hazan
2020 arXiv   pre-print
Empirically, we analyze the affect of this loss-gradient norm term using different deep nets.  ...  To avoid these assumptions, in this paper, we propose a new generalization bound for Bayesian deep nets by exploiting the contractivity of the Log-Sobolev inequalities.  ...  Unfortunately, this assumption excludes plenty of deep nets and Bayesian deep nets that rely on the popular negative log-likelihood (NLL) loss.  ... 
arXiv:2002.09866v1 fatcat:z7uuysgjavduzfcfcu43izjxli

Deep Learning in Multi-Class Lung Diseases' Classification on Chest X-ray Images

Sungyeup Kim, Beanbonyka Rim, Seongjun Choi, Ahyoung Lee, Sedong Min, Min Hong
2022 Diagnostics  
We experimented using our proposed method on three classes of normal, pneumonia, and pneumothorax of the U.S.  ...  We also experimented on the Cheonan Soonchunhyang University Hospital (SCH) data set on four classes of normal, pneumonia, pneumothorax, and tuberculosis, and achieved validation performances of loss =  ...  multi-class classification on the NIH data set (%).  ... 
doi:10.3390/diagnostics12040915 pmid:35453963 pmcid:PMC9025806 fatcat:mnkpuqtnvjhpppzts4543v7gmy

A Pornographic Images Recognition Model based on Deep One-Class Classification With Visual Attention Mechanism

Junren Chen, Gang Liang, Wenbo He, Chun Xu, Jin Yang, Ruihang Liu
2020 IEEE Access  
INDEX TERMS Pornography classification, deep learning, one-class classification, visual attention mechanism, adversarial attacks.  ...  In order to address this challenge, this paper proposes a method named Deep One-Class with Attention for Pornography (DOCAPorn) that recognizes the pornographic images through the one-class classification  ...  In order to improve the performance of deep one-class classification, the visual attention mechanism is introduced into the proposed deep one-class classification model.  ... 
doi:10.1109/access.2020.2988736 fatcat:xe6k6j2pxbeo5exycel3tzfelq

One-class Text Classification with Multi-modal Deep Support Vector Data Description

Chenlong Hu, Yukun Feng, Hidetaka Kamigaito, Hiroya Takamura, Manabu Okumura
2021 Journal of Natural Language Processing  
This work presents multi-modal deep SVDD (mSVDD) for one-class text classification.  ...  By extending the uni-modal SVDD to a multiple modal one, we build mSVDD with multiple hyperspheres, that enable us to build a much better description for target one-class data.  ...  More experiments and discussions on the incorporation of negative supervision are also included. Section 6 is newly added for the summary of related work.  ... 
doi:10.5715/jnlp.28.1053 fatcat:t5vz4wq5h5adhlxhsq6zise57m

Deep Decision Network for Multi-class Image Classification

Venkatesh N. Murthy, Vivek Singh, Terrence Chen, R. Manmatha, Dorin Comaniciu
2016 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR)  
We validate DDN on two publicly available benchmark datasets: CIFAR-10 and CIFAR-100 and it yields state-of-the-art classification performance on both the datasets.  ...  In this paper, we present a novel Deep Decision Network (DDN) that provides an alternative approach towards building an efficient deep learning network.  ...  Figure 3 . 3 DDN method idea validation on classification of digit '6' and '8' of MNIST dataset. left image indicates some of the confusion classes at level-1 and the right one indicates some confusion  ... 
doi:10.1109/cvpr.2016.246 dblp:conf/cvpr/MurthySCMC16 fatcat:hrjqqimf5fei7jauh4ix6dqwke
« Previous Showing results 1 — 15 out of 435,023 results