Filters








481,412 Hits in 5.7 sec

Label Error Correction and Generation through Label Relationships

Zijun Cui, Yong Zhang, Qiang Ji
2020 PROCEEDINGS OF THE THIRTIETH AAAI CONFERENCE ON ARTIFICIAL INTELLIGENCE AND THE TWENTY-EIGHTH INNOVATIVE APPLICATIONS OF ARTIFICIAL INTELLIGENCE CONFERENCE  
The relationships among labels, on other hand, are usually stable and robust to errors.  ...  For this reason, we propose to capture and leverage label relationships at different levels to improve fine-grained label annotation quality and to generate labels.  ...  Acknowledgment The work described in this paper is supported in part by a DARPA grant FA, and in part by the US National Science Foundation award CNS #1629856.  ... 
doi:10.1609/aaai.v34i04.5778 fatcat:svnhzpkuyzccxeil36jrwcobqq

A proposal for run-time checks on command execution order of software program

Shinichi Funase, Toshihiko Shimauchi, Haruhiko Kimura
2020 Studies in Science and Technology  
The proposed system detected 80% of the errors detected by the input / output relationship method.  ...  Therefore, the assembler language that has a one-to-one correspondence with the machine language and is expressed by alphanumeric characters and symbols is used for this paper.  ...  The general flowchart in Figure 1 shows a simple process that is divided into a process with label N 1 and that with label N2 at the branch point R 1 .  ... 
doi:10.11425/sst.9.149 fatcat:xad332k4svgwvme2yirfny7cma

An evaluation of labeling-then-doing with moderately handicapped persons: acquisition and generalization with complex tasks

D P Wacker, W K Berg, C McMahon, M Templeman, J McKinney, V Swarts, M Visser, P Marquardt
1988 Journal of Applied Behavior Analysis  
The performance of all students generalized across tasks and settings, and the use of labels generalized for 2 of the students.  ...  Following training, all subjects' use of verbal labels and keyentry skills generalized across tasks (programs) and settings (offices and computer terminals).  ...  Correction was provided for errors, and praise was provided for correct performance. Following a labeling error, the client was asked to repeat the label.  ... 
doi:10.1901/jaba.1988.21-369 pmid:3225254 pmcid:PMC1286136 fatcat:vlzkalwk7vat3kln5vd5yaruxy

Effective Building Extraction by Learning to Detect and Correct Erroneous Labels in Segmentation Mask

Praveer Singh, Nikos Komodakis
2018 IGARSS 2018 - 2018 IEEE International Geoscience and Remote Sensing Symposium  
These Error maps are then used to correct the corresponding erroneous labels through a replacement technique.  ...  One of the primary reasons is that they partially or wholly, neglect the underlying relationship that exist in the joint space of input and output variables.  ...  However, these refinements through residual correction can only correct small errors (such as on the boundaries) while leaving major segmentation errors intact.  ... 
doi:10.1109/igarss.2018.8517854 dblp:conf/igarss/SinghK18 fatcat:eo7gopjldjbxlhbav5ilr4onky

Toddlers benefit from labeling on an executive function search task

Stephanie E. Miller, Stuart Marcovitch
2011 Journal of Experimental Child Psychology  
The results revealed that accuracy improved across conditions such that children made fewest errors when they generated the label for the hiding location.  ...  and labeled by the experimenter, or (d) marked by a familiar picture and labeled by the participant.  ...  Further, generation should be especially important to young children who are just beginning to appreciate dual representations through self-initiated pointing and labeling.  ... 
doi:10.1016/j.jecp.2010.10.008 pmid:21112597 pmcid:PMC3042530 fatcat:2alv77h4qze25l7zml4s5jwupy

Software Engineering for Spreadsheets

Martin Erwig
2009 IEEE Software  
Since many critical decisions are made based on values computed by spreadsheets, the correctness of spreadsheets is crucial.  ...  The application of tried and tested software engineering principles to spreadsheets seems to promise help with the construction and maintenance of dependable spreadsheets.  ...  For example, B3 is labeled by "Month" and "Apple". • The labeling relationship exhibits, in general, a hierarchical structure.  ... 
doi:10.1109/ms.2009.140 fatcat:xbqrgisjbzghdcokvmafbbc3lu

A Knowledge Based Approach for Tackling Mislabeled Multi-class Big Social Data [chapter]

Minyi Guo, Yi Liu, Jie Li, Huakang Li, Bei Xu
2014 Lecture Notes in Computer Science  
the error labels in big data.  ...  In this paper, we propose a knowledge based approach for tackling mislabeled multi-class big data, in which knowledge graph technique is combined with other data correction method to perceive and correct  ...  The polishing algorithm can predict and correct both attributes errors and label errors(i.e. class errors). In this paper, we use it to correct label errors.  ... 
doi:10.1007/978-3-319-07443-6_24 fatcat:lqn3gh7n4rbxhp6y4djoa6dilm

Application and Research of Active RFID-Based Positioning System in Sport Competition

Jing Ning
2016 International Journal of Future Generation Communication and Networking  
Through the improved location algorithm, it builds a relationship between the distance and time in finish line of the athlete, and solves the function to make sure the time that athletes reach the finish  ...  In order to reduce the timing error, we can bind some RFID labels on the shoes of athlete. These RFID labels do timing at the same time, and we select the average.  ...  Through positioning algorithm, form corresponding function relationship between distance and time from the finish line.  ... 
doi:10.14257/ijfgcn.2016.9.12.16 fatcat:ap22om34i5asbiuaxbvlrlojuy

Improving quality of training data for learning to rank using click-through data

Jingfang Xu, Chuanliang Chen, Gu Xu, Hang Li, Elbio Renato Torres Abib
2010 Proceedings of the third ACM international conference on Web search and data mining - WSDM '10  
Finally, we verify that using training data in which the errors are detected and corrected by our method we can improve the performance of learning to rank algorithms.  ...  Secondly, we propose detecting relevance judgment errors using click-through data accumulated at a search engine.  ...  [1] worked on automatic generation of training data from click-through data.  ... 
doi:10.1145/1718487.1718509 dblp:conf/wsdm/XuCXLA10 fatcat:ez2g4b3wwjdy7iebuxtkzf3dpi

Active Image Labeling and Its Application to Facial Action Labeling [chapter]

Lei Zhang, Yan Tong, Qiang Ji
2008 Lecture Notes in Computer Science  
To minimize the human involvement, an active user feedback strategy is developed, through which the optimal user feedback is determined, so that the labeling errors in the subsequent re-labeling process  ...  Manual data labeling is labor-intensive and prone to the human errors. The training data it produces often lacks in both quantity and quality.  ...  [29] , there are semantic relationships (the co-occurrence relationships and the mutually exclusive relationships) among AUs.  ... 
doi:10.1007/978-3-540-88688-4_52 fatcat:dvxqeeji3ba57ivlq6nyjjwrhe

Generating Conceptual Subgraph from Tabular Data for Knowledge Graph Matching

Donguk Kim, Heesung Park, Jae Kyu Lee, Wooju Kim
2020 International Semantic Web Conference  
It is most important to find the label that indicates the correct meaning in Wikidata where data and values are annotated with each label.  ...  Wikidata has a label for each document. In addition, it has the characteristic of being linked to another document through these documents. These connected data can be represented as graphs.  ...  in Table 2 and generate a conceptual subgraph as shown in Figure 2 .  ... 
dblp:conf/semweb/KimPLK20 fatcat:jushz4oo5fbgdnufiej6aisihy

A Kind of Intelligent Substation Layer Fiber Smart Tag Technology Research

Ning YU, Feng WANG
2017 DEStech Transactions on Computer Science and Engineering  
Analyzes the cable information "nakedness corresponding" principle, this paper proposes a smart tag technology, based on the QR code mode and the coding, generation of smart labels and parsing is studied  ...  Results show that the technology can greatly improve the efficiency of the intelligent substation debugging, maintenance and reconstruction and correctness.  ...  technology, the research process of intelligent substation layer fiber smart tag coding, generating and analytic method, implementation process layer cable label information "that is sweeping the look"  ... 
doi:10.12783/dtcse/cece2017/14366 fatcat:vbntz5p42zhfdop7s2ufgynnym

Evaluating structural pattern recognition for handwritten math via primitive label graphs

Richard Zanibbi, Harold Mouchère, Christian Viard-Gaudin, Richard Zanibbi, Bertrand Coüasnon
2013 Document Recognition and Retrieval XX  
classification and relationships.  ...  We define new metrics obtained by Hamming distances over label graphs, which allow classification, segmentation and parsing errors to be characterized separately, or using a single measure.  ...  ACKNOWLEDGMENTS Our thanks to Bertrand Coüasnon and Aurélie Lemaitre and the anonymous reviewers for helpful feedback, and to the CROHME 2012 participants for sharing their system outputs.  ... 
doi:10.1117/12.2008409 dblp:conf/drr/ZanibbiMV13 fatcat:gd7gdab5ive6db7xm4rjxxrj7q

Detecting Label Errors using Pre-Trained Language Models [article]

Derek Chong, Jenny Hong, Christopher D. Manning
2022 arXiv   pre-print
outperforms more complex mechanisms for detecting label errors on natural language datasets.  ...  We contribute a novel method to produce highly realistic, human-originated label noise from crowdsourced data, and demonstrate the effectiveness of this method on TweetNLP, providing an otherwise difficult  ...  Through cross-model experiments on TweetNLP-5, we observe a relationship between a model's ability to detect label errors and its performance on a standard natural language undersatnding (NLU) benchmark  ... 
arXiv:2205.12702v1 fatcat:a32b3qldx5a7zeytewieipewoi

Online Reasoning for Ontology-Based Error Detection in Text [chapter]

Fernando Gutiererz, Dejing Dou, Stephen Fickas, Gina Griffiths
2014 Lecture Notes in Computer Science  
., property domain and range) likewise affect the capability of detecting errors.  ...  In this paper we propose a new approach that uses logic reasoning to detect errors in a statement from text online.  ...  The views and conclusions contained in this document are those of the authors and should not be interpreted as necessarily representing the official policies, either expressed or implied, of the NSF.  ... 
doi:10.1007/978-3-662-45563-0_34 fatcat:nrpdktejyjasfiol7qv3nvja3i
« Previous Showing results 1 — 15 out of 481,412 results