this paper examines the human expressive states dependent on facial pictures utilizing a few viable component extraction methods. It reproduces the K-Nearest Neighbor (k-NN) classifier to approve the adequacy of succe...
详细信息
this paper examines the human expressive states dependent on facial pictures utilizing a few viable component extraction methods. It reproduces the K-Nearest Neighbor (k-NN) classifier to approve the adequacy of successful capabilities separated from the Local Binary pattern (LBP) and Histograms of Oriented Gradients (HOG) for the said task. An examination of the strategies has been made dependent on the normal acknowledgment precision of the classifiers utilizing the calculation unpredictability as a compromise. the component extraction methods have been approved for their discriminative force under various preparations for testing information division proportions, Kappa Coefficient, and order time. the LBP has outperformed the HOG include extraction strategy with a normal precision of 79.6% yet remains computationally costly. On the contrary, the HOG method has furnished a lower characterization time with a normal precision of 59.3 % as uncovered from our outcomes.
Table detection and structure recognition from archival document images remain challenging due to diverse table structures, complex document layouts, degraded image qualities and inconsistent table scales. In this pap...
详细信息
ISBN:
(纸本)9783030865498
Table detection and structure recognition from archival document images remain challenging due to diverse table structures, complex document layouts, degraded image qualities and inconsistent table scales. In this paper, we propose an instance segmentation based approach for archival table structure recognition which utilizes both foreground cell content and background ruling line information. To overcome the influence from inconsistent table scales, we design an adaptive image scaling method based on average cell size and density of ruling lines inside each document image. Different from previous multi-scale training and testing approaches which usually slow down the speed of the whole system, our adaptive scaling resizes each image to a single optimal size which can not only improve overall model performance but also reduce memory and computing overhead on average. Extensive experiments on cTDaR 2019 Archival dataset show that our method can outperform the baselines and achieve new state-of-the-art performance, which demonstrates the effectiveness and superiority of the proposed method.
In this paper, a patternrecognition algorithm is given based on centroids of fuzzy hyper-pyramid numbers which are special type fuzzy n-cell numbers. the specific calculation formula (which can be easy calculated by ...
详细信息
the proceedings contain 71 papers. the topics discussed include: implementation of machine learning to detect hate speech in Bangla language;analyzing recent research trends of computer science from academic open-acce...
ISBN:
(纸本)9781728132457
the proceedings contain 71 papers. the topics discussed include: implementation of machine learning to detect hate speech in Bangla language;analyzing recent research trends of computer science from academic open-access digital library;simulation and modeling of seawater intrusion around Pondicherry coastal aquifer — India;an overview of fog computing in the present scenario;text extraction through video lip reading using deep learning;multiservice online platform for integrated geospatial data processing;generating Bengali news headlines: an attentive approach with sequence-to-sequence networks;and periodic monitoring of rivers using portable sensor system.
this paper presents a 'Unified Side Channel Attack Model' (USCA-M). the USCA-M model is compiled by the research undertaken of side-channel attacks (SCAs) from published journal articles and conference papers ...
详细信息
the proceedings contain 39 papers. the topics discussed include: (dis)similarity-based correlation functions on polar scales;logic of estimates for fuzzy logics sentences;architecture of some models for optimization p...
the proceedings contain 39 papers. the topics discussed include: (dis)similarity-based correlation functions on polar scales;logic of estimates for fuzzy logics sentences;architecture of some models for optimization problems under conditions of hybrid uncertainty;inference method for MISO-structure systems with fuzzy inputs using parallel computing techniques;method for classification of objects with fuzzy values of features;analysis of the multifactorial phenomena based on fuzzy Bayesian model;fuzzy clustering in the problem of technological state assessment for phosporose production;algebraic Bayesian networks: a frequentist approach to knowledge pattern parameters machine learning;empirical approach to assessing sensitivity of local posteriori inference of algebraic Bayesian network;and new trends in measurement theory: Bayesian intelligent measurement and its application in the digital economy.
the proceedings contain 37 papers. the special focus in this conference is on internationalconference on Analysis of Images, Social Networks and Texts. the topics include: Automated approach to rhythm figures search ...
ISBN:
(纸本)9783030395742
the proceedings contain 37 papers. the special focus in this conference is on internationalconference on Analysis of Images, Social Networks and Texts. the topics include: Automated approach to rhythm figures search in english text;Adapting the Graph2Vec approach to dependency trees for NLP tasks;recognition of parts of speech using the vector of bigram frequencies;formalization of medical records using an ontology: Patient complaints;a deep learning method study of user interest classification;morpheme segmentation for russian: Evaluation of convolutional neural network models;Using pre-trained deeply contextual model BERT for Russian named entity recognition;expert assessment of synonymic rows in ruwordnet;Text mining for evaluation of candidates based on their CVs;evolutionary algorithms for constructing an ensemble of decision trees;vec2graph: A python library for visualizing word embeddings as graphs;how to prevent harmful information spreading in social networks using simulation tools;effect of social graph structure on the utilization rate in a flat organization;discernment of smoke in image frames;Face recognition using DELF feature descriptors on RGB-D data;efficient information support of the automatic process and production control system;an ensemble of learning machine models for plant recognition;hand gestures detection, tracking and classification using convolutional neural network;synthesizing data using variational autoencoders for handling class imbalanced deep learning;new approach for fast residual strain estimation through rational 2D diffraction pattern processing;an algorithm for constructing a topological skeleton for semi-structured spatial data based on persistent homology;fast identification of fingerprint.
Deep learning is a machine learning technique that is inspired by human brain. It enables information processing in multiple hierarchical layers to understand representations and features from raw data. Deep learning ...
详细信息
Regardless of different word embedding and hidden layer structures of the neural architectures that are used in named entity recognition, a conditional random field layer is commonly used for the output. this work pro...
详细信息
the article presents the results of the study of the upper assessment of the sensitivity of solving the second problem of local posterior inference in algebraic Bayesian networks to variations in probability estimates...
详细信息
暂无评论