Visual Explanations via Iterated Integrated Attributions

We introduce Iterated Integrated Attributions (IIA) - a generic method for explaining the predictions of vision models. IIA employs iterative integration across the input image, the internal representations generated by the model, and their gradients, yielding precise and focused explanation maps. W...

Full description

Saved in:
Bibliographic Details
Published in:arXiv.org 2023-10
Main Authors: Barkan, Oren, Yehonatan Elisha, Asher, Yuval, Eshel, Amit, Koenigstein, Noam
Format: Article
Language:English
Subjects:
Iterative methods
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
cited_by
cites
container_end_page
container_issue
container_start_page
container_title arXiv.org
container_volume
creator Barkan, Oren
Yehonatan Elisha
Asher, Yuval
Eshel, Amit
Koenigstein, Noam
description We introduce Iterated Integrated Attributions (IIA) - a generic method for explaining the predictions of vision models. IIA employs iterative integration across the input image, the internal representations generated by the model, and their gradients, yielding precise and focused explanation maps. We demonstrate the effectiveness of IIA through comprehensive evaluations across various tasks, datasets, and network architectures. Our results showcase that IIA produces accurate explanation maps, outperforming other state-of-the-art explanation techniques.
format article
fullrecord <record><control><sourceid>proquest</sourceid><recordid>TN_cdi_proquest_journals_2884474344</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><sourcerecordid>2884474344</sourcerecordid><originalsourceid>FETCH-proquest_journals_28844743443</originalsourceid><addsrcrecordid>eNpjYuA0MjY21LUwMTLiYOAtLs4yMDAwMjM3MjU15mSwCMssLk3MUXCtKMhJzEssyczPK1Yoy0xU8CxJLUosSU1R8MwrSU2HMB1LSooyk0rBingYWNMSc4pTeaE0N4Oym2uIs4duQVF-YWlqcUl8Vn5pUR5QKt7IwsLExNzE2MTEmDhVAIDXNqg</addsrcrecordid><sourcetype>Aggregation Database</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype><pqid>2884474344</pqid></control><display><type>article</type><title>Visual Explanations via Iterated Integrated Attributions</title><source>Publicly Available Content Database (Proquest) (PQ_SDU_P3)</source><creator>Barkan, Oren ; Yehonatan Elisha ; Asher, Yuval ; Eshel, Amit ; Koenigstein, Noam</creator><creatorcontrib>Barkan, Oren ; Yehonatan Elisha ; Asher, Yuval ; Eshel, Amit ; Koenigstein, Noam</creatorcontrib><description>We introduce Iterated Integrated Attributions (IIA) - a generic method for explaining the predictions of vision models. IIA employs iterative integration across the input image, the internal representations generated by the model, and their gradients, yielding precise and focused explanation maps. We demonstrate the effectiveness of IIA through comprehensive evaluations across various tasks, datasets, and network architectures. Our results showcase that IIA produces accurate explanation maps, outperforming other state-of-the-art explanation techniques.</description><identifier>EISSN: 2331-8422</identifier><language>eng</language><publisher>Ithaca: Cornell University Library, arXiv.org</publisher><subject>Iterative methods</subject><ispartof>arXiv.org, 2023-10</ispartof><rights>2023. This work is published under http://arxiv.org/licenses/nonexclusive-distrib/1.0/ (the “License”). Notwithstanding the ProQuest Terms and Conditions, you may use this content in accordance with the terms of the License.</rights><oa>free_for_read</oa><woscitedreferencessubscribed>false</woscitedreferencessubscribed></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><linktohtml>$$Uhttps://www.proquest.com/docview/2884474344?pq-origsite=primo$$EHTML$$P50$$Gproquest$$Hfree_for_read</linktohtml><link.rule.ids>780,784,25753,37012,44590</link.rule.ids></links><search><creatorcontrib>Barkan, Oren</creatorcontrib><creatorcontrib>Yehonatan Elisha</creatorcontrib><creatorcontrib>Asher, Yuval</creatorcontrib><creatorcontrib>Eshel, Amit</creatorcontrib><creatorcontrib>Koenigstein, Noam</creatorcontrib><title>Visual Explanations via Iterated Integrated Attributions</title><title>arXiv.org</title><description>We introduce Iterated Integrated Attributions (IIA) - a generic method for explaining the predictions of vision models. IIA employs iterative integration across the input image, the internal representations generated by the model, and their gradients, yielding precise and focused explanation maps. We demonstrate the effectiveness of IIA through comprehensive evaluations across various tasks, datasets, and network architectures. Our results showcase that IIA produces accurate explanation maps, outperforming other state-of-the-art explanation techniques.</description><subject>Iterative methods</subject><issn>2331-8422</issn><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2023</creationdate><recordtype>article</recordtype><sourceid>PIMPY</sourceid><recordid>eNpjYuA0MjY21LUwMTLiYOAtLs4yMDAwMjM3MjU15mSwCMssLk3MUXCtKMhJzEssyczPK1Yoy0xU8CxJLUosSU1R8MwrSU2HMB1LSooyk0rBingYWNMSc4pTeaE0N4Oym2uIs4duQVF-YWlqcUl8Vn5pUR5QKt7IwsLExNzE2MTEmDhVAIDXNqg</recordid><startdate>20231028</startdate><enddate>20231028</enddate><creator>Barkan, Oren</creator><creator>Yehonatan Elisha</creator><creator>Asher, Yuval</creator><creator>Eshel, Amit</creator><creator>Koenigstein, Noam</creator><general>Cornell University Library, arXiv.org</general><scope>8FE</scope><scope>8FG</scope><scope>ABJCF</scope><scope>ABUWG</scope><scope>AFKRA</scope><scope>AZQEC</scope><scope>BENPR</scope><scope>BGLVJ</scope><scope>CCPQU</scope><scope>DWQXO</scope><scope>HCIFZ</scope><scope>L6V</scope><scope>M7S</scope><scope>PIMPY</scope><scope>PQEST</scope><scope>PQQKQ</scope><scope>PQUKI</scope><scope>PRINS</scope><scope>PTHSS</scope></search><sort><creationdate>20231028</creationdate><title>Visual Explanations via Iterated Integrated Attributions</title><author>Barkan, Oren ; Yehonatan Elisha ; Asher, Yuval ; Eshel, Amit ; Koenigstein, Noam</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-proquest_journals_28844743443</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2023</creationdate><topic>Iterative methods</topic><toplevel>online_resources</toplevel><creatorcontrib>Barkan, Oren</creatorcontrib><creatorcontrib>Yehonatan Elisha</creatorcontrib><creatorcontrib>Asher, Yuval</creatorcontrib><creatorcontrib>Eshel, Amit</creatorcontrib><creatorcontrib>Koenigstein, Noam</creatorcontrib><collection>ProQuest SciTech Collection</collection><collection>ProQuest Technology Collection</collection><collection>Materials Science &amp; Engineering Collection</collection><collection>ProQuest Central (Alumni)</collection><collection>ProQuest Central</collection><collection>ProQuest Central Essentials</collection><collection>ProQuest Central</collection><collection>Technology Collection</collection><collection>ProQuest One Community College</collection><collection>ProQuest Central</collection><collection>SciTech Premium Collection</collection><collection>ProQuest Engineering Collection</collection><collection>Engineering Database</collection><collection>Publicly Available Content Database (Proquest) (PQ_SDU_P3)</collection><collection>ProQuest One Academic Eastern Edition (DO NOT USE)</collection><collection>ProQuest One Academic</collection><collection>ProQuest One Academic UKI Edition</collection><collection>ProQuest Central China</collection><collection>Engineering collection</collection></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext</fulltext></delivery><addata><au>Barkan, Oren</au><au>Yehonatan Elisha</au><au>Asher, Yuval</au><au>Eshel, Amit</au><au>Koenigstein, Noam</au><format>book</format><genre>document</genre><ristype>GEN</ristype><atitle>Visual Explanations via Iterated Integrated Attributions</atitle><jtitle>arXiv.org</jtitle><date>2023-10-28</date><risdate>2023</risdate><eissn>2331-8422</eissn><abstract>We introduce Iterated Integrated Attributions (IIA) - a generic method for explaining the predictions of vision models. IIA employs iterative integration across the input image, the internal representations generated by the model, and their gradients, yielding precise and focused explanation maps. We demonstrate the effectiveness of IIA through comprehensive evaluations across various tasks, datasets, and network architectures. Our results showcase that IIA produces accurate explanation maps, outperforming other state-of-the-art explanation techniques.</abstract><cop>Ithaca</cop><pub>Cornell University Library, arXiv.org</pub><oa>free_for_read</oa></addata></record>
fulltext fulltext
identifier EISSN: 2331-8422
ispartof arXiv.org, 2023-10
issn 2331-8422
language eng
recordid cdi_proquest_journals_2884474344
source Publicly Available Content Database (Proquest) (PQ_SDU_P3)
subjects Iterative methods
title Visual Explanations via Iterated Integrated Attributions
url http://sfxeu10.hosted.exlibrisgroup.com/loughborough?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2024-12-21T15%3A55%3A49IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-proquest&rft_val_fmt=info:ofi/fmt:kev:mtx:book&rft.genre=document&rft.atitle=Visual%20Explanations%20via%20Iterated%20Integrated%20Attributions&rft.jtitle=arXiv.org&rft.au=Barkan,%20Oren&rft.date=2023-10-28&rft.eissn=2331-8422&rft_id=info:doi/&rft_dat=%3Cproquest%3E2884474344%3C/proquest%3E%3Cgrp_id%3Ecdi_FETCH-proquest_journals_28844743443%3C/grp_id%3E%3Coa%3E%3C/oa%3E%3Curl%3E%3C/url%3E&rft_id=info:oai/&rft_pqid=2884474344&rft_id=info:pmid/&rfr_iscdi=true