统一的可解释AI框架，提高分类器性能

2019 Grace Hopper Celebration India (GHCI) Pub Date : 2019-11-01 DOI:10.1109/GHCI47972.2019.9071811

R. Manjunath, B.N Chandrashekar, B. Vinutha, Rahul Arya, Arindam Chatterjee

{"title":"统一的可解释AI框架，提高分类器性能","authors":"R. Manjunath, B.N Chandrashekar, B. Vinutha, Rahul Arya, Arindam Chatterjee","doi":"10.1109/GHCI47972.2019.9071811","DOIUrl":null,"url":null,"abstract":"Deep learning image classifiers are extensively used in document processing, activity monitoring, object recognition and separations etc. However, even the best classifiers are not free from errors. It would be very helpful if the errors that are pumped in to the system due to the classifier decisions are reduced. The framework comprises of heat map generation, attribute generation, text explanation generation and activation.","PeriodicalId":153240,"journal":{"name":"2019 Grace Hopper Celebration India (GHCI)","volume":"48 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2019-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"Unified framework of Explainable AI to enhance classifier performance\",\"authors\":\"R. Manjunath, B.N Chandrashekar, B. Vinutha, Rahul Arya, Arindam Chatterjee\",\"doi\":\"10.1109/GHCI47972.2019.9071811\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"Deep learning image classifiers are extensively used in document processing, activity monitoring, object recognition and separations etc. However, even the best classifiers are not free from errors. It would be very helpful if the errors that are pumped in to the system due to the classifier decisions are reduced. The framework comprises of heat map generation, attribute generation, text explanation generation and activation.\",\"PeriodicalId\":153240,\"journal\":{\"name\":\"2019 Grace Hopper Celebration India (GHCI)\",\"volume\":\"48 1\",\"pages\":\"0\"},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2019-11-01\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"2019 Grace Hopper Celebration India (GHCI)\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/10.1109/GHCI47972.2019.9071811\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"2019 Grace Hopper Celebration India (GHCI)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/GHCI47972.2019.9071811","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}

引用次数: 0

摘要

深度学习图像分类器广泛应用于文档处理、活动监控、目标识别和分离等领域。然而，即使是最好的分类器也不是没有错误的。如果由于分类器决策而被泵入系统的错误减少，这将非常有帮助。该框架包括热图生成、属性生成、文本解释生成和激活。

本文章由计算机程序翻译，如有差异，请以英文原文为准。

查看原文本刊更多论文

Unified framework of Explainable AI to enhance classifier performance

Deep learning image classifiers are extensively used in document processing, activity monitoring, object recognition and separations etc. However, even the best classifiers are not free from errors. It would be very helpful if the errors that are pumped in to the system due to the classifier decisions are reduced. The framework comprises of heat map generation, attribute generation, text explanation generation and activation.

求助全文

通过发布文献求助，成功后即可免费获取论文全文。去求助

来源期刊

2019 Grace Hopper Celebration India (GHCI)

自引率

0.00%

发文量