Prototype-Based Explainable Deep Learning for SexClassification of Mountain Gazelles in the Wild

Prototype-Based Explainable Deep Learning for SexClassification of Mountain Gazelles in the Wild | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Prototype-Based Explainable Deep Learning for SexClassification of Mountain Gazelles in the Wild ‪Tali Boneh Shitrit‬‏, Amir Kedem, Efrat Yagur, Ilan Shimshomi, and 1 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-8659636/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted 10 You are reading this latest preprint version Abstract Wildlife preserves are critical for countering the decline of biodiversity, yet monitoring species distribution and demographics viadirect observation is time-consuming and prone to observer disturbance. Deep learning has emerged as a scalable alternativefor analyzing field data, yet the ’black-box’ nature of standard models hinders their adoption. This is a major obstacle inecological research, where transparent reasoning is essential for trust and meaningful interpretation. We address this challenge through the lens of sex classification in the endangered Mountain Gazelle (Gazella gazella). Reliableautomated classification from field imagery is essential for assessing population structure and welfare metrics for effectiveconservation management. While Mountain Gazelles exhibit distinct sexual dimorphism in horn morphology and body size,inferring sex from unconstrained camera trap data remains challenging due to significant variability in pose, illumination, and occlusion. To enhance explainability in Mountain Gazelle sex classification, we employed a prototype-based deep learning framework.Unlike standard “black-box” models, this approach represents classes through learned visual exemplars (prototypes) thatcorrespond directly to crucial image features, ensuring transparent reasoning. Building upon the PIP-Net architecture, weintroduced a novel enhancement: the ability to learn prototypes of variable sizes and aspect ratios, moving beyond rigid,fixed-size patches. These prototypes are learned autonomously without manual trait annotation, allowing the model to discoverbiologically meaningful regions such as horn morphology, body width, and leg configuration. By optimizing across a listof candidate scales, our non-fixed prototype approach captures larger, semantically coherent regions, leading to improvedpredictive performance and detailed explanations. Our analysis demonstrates that the most influential prototypes align withecological knowledge: both sexes are identified via central body, leg, and head features , with the model distinguishing betweenthem based on sex-specific morphological proportions. For males, horn-related prototypes serve as a distinct cue, reflectingthe main dimorphic trait cited in literature. These representations confirm that the models decision-making is grounded ingenuine morphological traits rather than background artifacts, offering a transparent and accurate tool for wildlife classification. Biological sciences/Computational biology and bioinformatics Biological sciences/Ecology Earth and environmental sciences/Ecology Biological sciences/Evolution Biological sciences/Zoology Full Text Additional Declarations No competing interests reported. Cite Share Download PDF Status: Under Review Version 1 posted Reviewers agreed at journal 16 May, 2026 Reviewers agreed at journal 15 May, 2026 Reviews received at journal 28 Apr, 2026 Reviewers agreed at journal 27 Apr, 2026 Reviewers agreed at journal 23 Apr, 2026 Reviewers invited by journal 22 Apr, 2026 Editor invited by journal 29 Jan, 2026 Editor assigned by journal 24 Jan, 2026 Submission checks completed at journal 24 Jan, 2026 First submitted to journal 21 Jan, 2026 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-8659636","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":628080822,"identity":"db697696-1b72-4454-b3b5-c835bb677e1e","order_by":0,"name":"‪Tali Boneh Shitrit‬‏","email":"","orcid":"","institution":"University of Haifa","correspondingAuthor":false,"prefix":"","firstName":"‪Tali","middleName":"Boneh","lastName":"Shitrit‬‏","suffix":""},{"id":628080824,"identity":"5ae2e951-67e0-40e5-a1a8-4636ed197fdc","order_by":1,"name":"Amir Kedem","email":"","orcid":"","institution":"University of Haifa","correspondingAuthor":false,"prefix":"","firstName":"Amir","middleName":"","lastName":"Kedem","suffix":""},{"id":628080829,"identity":"3f1eeb50-30e6-4dd8-a915-faf05cf24123","order_by":2,"name":"Efrat Yagur","email":"","orcid":"","institution":"Jack Joseph and Morton Mandel Gazelle Valley","correspondingAuthor":false,"prefix":"","firstName":"Efrat","middleName":"","lastName":"Yagur","suffix":""},{"id":628080831,"identity":"40254a88-97b3-4019-b4b2-7eb1f6cb5bcb","order_by":3,"name":"Ilan Shimshomi","email":"","orcid":"","institution":"University of Haifa","correspondingAuthor":false,"prefix":"","firstName":"Ilan","middleName":"","lastName":"Shimshomi","suffix":""},{"id":628080833,"identity":"1b5ea409-bf7d-4ad6-b842-e75c0b6fe640","order_by":4,"name":"Anna Zamansky","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA8UlEQVRIiWNgGAWjYBADOTYGBiAykINwEwoIazGGajGGajEgrCWxAayFAaqFAY8W/tm9Bz/+qNmW3sd/gO3BjwIDeQb2ww8YHuDRInHnXLKExLHbuW0SCeyGPQYGhg08aQb4HXYjx0DCgA2khYFNgsfgD2MDQw5+v8jfyDH+kfDvdjob0GGSfwwM7Bv43+DXYnAjx0ziYNvtBDaGBDZpHgODxAYJArYYArVYNvbdNmyTSGyTljEwSG6TeGZwAJ8WOaDDbv74dltevv/wMck3fwxs+/mTHz78UYHH+wgA9DgIgGLnAFEaRsEoGAWjYBTgBACZjEjBczyuWAAAAABJRU5ErkJggg==","orcid":"","institution":"University of Haifa","correspondingAuthor":true,"prefix":"","firstName":"Anna","middleName":"","lastName":"Zamansky","suffix":""}],"badges":[],"createdAt":"2026-01-21 12:09:10","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-8659636/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-8659636/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":107707809,"identity":"81a194ff-5ab7-4e42-94be-25ca6b12b691","added_by":"auto","created_at":"2026-04-24 09:21:10","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1042669,"visible":true,"origin":"","legend":"","description":"","filename":"TaliExplainabilityGazelles3.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8659636/v1_covered_1a8e7c6b-5f96-4256-865b-4c7a745a7454.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Prototype-Based Explainable Deep Learning for SexClassification of Mountain Gazelles in the Wild","fulltext":[],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":false,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":true,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":true,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"scientific-reports","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"scirep","sideBox":"Learn more about [Scientific Reports](http://www.nature.com/srep/)","snPcode":"","submissionUrl":"","title":"Scientific Reports","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Scientific Reports","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-8659636/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-8659636/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eWildlife preserves are critical for countering the decline of biodiversity, yet monitoring species distribution and demographics viadirect observation is time-consuming and prone to observer disturbance. Deep learning has emerged as a scalable alternativefor analyzing field data, yet the ’black-box’ nature of standard models hinders their adoption. This is a major obstacle inecological research, where transparent reasoning is essential for trust and meaningful interpretation.\u003c/p\u003e\n\u003cp\u003eWe address this challenge through the lens of sex classification in the endangered Mountain Gazelle (Gazella gazella). Reliableautomated classification from field imagery is essential for assessing population structure and welfare metrics for effectiveconservation management. While Mountain Gazelles exhibit distinct sexual dimorphism in horn morphology and body size,inferring sex from unconstrained camera trap data remains challenging due to significant variability in pose, illumination, and occlusion.\u003c/p\u003e\n\u003cp\u003eTo enhance explainability in Mountain Gazelle sex classification, we employed a prototype-based deep learning framework.Unlike standard “black-box” models, this approach represents classes through learned visual exemplars (prototypes) thatcorrespond directly to crucial image features, ensuring transparent reasoning. Building upon the PIP-Net architecture, weintroduced a novel enhancement: the ability to learn prototypes of variable sizes and aspect ratios, moving beyond rigid,fixed-size patches. These prototypes are learned autonomously without manual trait annotation, allowing the model to discoverbiologically meaningful regions such as horn morphology, body width, and leg configuration. By optimizing across a listof candidate scales, our non-fixed prototype approach captures larger, semantically coherent regions, leading to improvedpredictive performance and detailed explanations. Our analysis demonstrates that the most influential prototypes align withecological knowledge: both sexes are identified via central body, leg, and head features , with the model distinguishing betweenthem based on sex-specific morphological proportions. For males, horn-related prototypes serve as a distinct cue, reflectingthe main dimorphic trait cited in literature. These representations confirm that the models decision-making is grounded ingenuine morphological traits rather than background artifacts, offering a transparent and accurate tool for wildlife classification.\u003c/p\u003e","manuscriptTitle":"Prototype-Based Explainable Deep Learning for SexClassification of Mountain Gazelles in the Wild","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-04-24 04:17:13","doi":"10.21203/rs.3.rs-8659636/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"reviewerAgreed","content":"758680143132363912197061084068413512","date":"2026-05-16T13:04:17+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"41825050983209589315910985799621701647","date":"2026-05-15T16:10:32+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-04-28T09:17:31+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"250647314481194622031309198788850869891","date":"2026-04-27T16:59:42+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"179621969948571406911587575618405136587","date":"2026-04-23T18:09:38+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2026-04-22T10:17:23+00:00","index":"","fulltext":""},{"type":"editorInvited","content":"","date":"2026-01-29T06:24:38+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2026-01-24T06:26:17+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2026-01-24T06:24:39+00:00","index":"","fulltext":""},{"type":"submitted","content":"Scientific Reports","date":"2026-01-21T11:55:45+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"scientific-reports","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"scirep","sideBox":"Learn more about [Scientific Reports](http://www.nature.com/srep/)","snPcode":"","submissionUrl":"","title":"Scientific Reports","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Scientific Reports","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"af4a0205-67f9-4949-95b4-61fffbae5eef","owner":[],"postedDate":"April 24th, 2026","published":true,"recentEditorialEvents":[{"type":"reviewerAgreed","content":"758680143132363912197061084068413512","date":"2026-05-16T13:04:17+00:00","index":133,"fulltext":""},{"type":"reviewerAgreed","content":"41825050983209589315910985799621701647","date":"2026-05-15T16:10:32+00:00","index":129,"fulltext":""}],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[{"id":66851017,"name":"Biological sciences/Computational biology and bioinformatics"},{"id":66851018,"name":"Biological sciences/Ecology"},{"id":66851019,"name":"Earth and environmental sciences/Ecology"},{"id":66851020,"name":"Biological sciences/Evolution"},{"id":66851021,"name":"Biological sciences/Zoology"}],"tags":[],"updatedAt":"2026-04-24T04:17:13+00:00","versionOfRecord":[],"versionCreatedAt":"2026-04-24 04:17:13","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-8659636","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-8659636","identity":"rs-8659636","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}