{"paper_id":"004e2a62-8a4e-4e6b-b9b8-f1f30a5fc3d2","body_text":"Cosine Similarity of Convolutional Features of Video Frames as a Method for Finding Similar Activities from Past and Present Information of Video Input | 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 Method Article Cosine Similarity of Convolutional Features of Video Frames as a Method for Finding Similar Activities from Past and Present Information of Video Input Franz Chuquirachi, Chyan Zheng Siow, Wei Hong Chin, Naoyuki Kubota This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-6362240/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Amid the declining workforce in eldercare, researchers are expected to contribute with new technology and applications to improve productivity and reduce the complexity of the caregiver job. In the development of Monitoring Systems, the task of recognizing the need for help from images is vital. However, past solutions present some problems: solutions constrained to a list of cases, and dependency on complete recordings of activity. In this paper, we presented an original definition for a situation that requires help and proposed a method to solve the component related to finding repetitions of human activity, while overcoming the mentioned issues. We used the cosine similarity between convolutional features of frames of a video to find similar activity using only past and present information. We proved the effectiveness of the method by challenging it to count the number of repetitions of an activity in a video and comparing the results with the ground truth values. The resulting accuracy was 63.63%, showing signs that it can be used for constructing generalized solutions with more immediate output delivery. Artificial Intelligence and Machine Learning Eldercare Monitoring Systems Activity Recognition Human Activity Repetition Cosine Similarity Convolutional Neural Network Video Analysis Caregiver Assistance Machine Learning Full Text Additional Declarations The authors declare no competing interests. Cite Share Download PDF Status: Posted Version 1 posted 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-6362240\",\"acceptedTermsAndConditions\":true,\"allowDirectSubmit\":true,\"archivedVersions\":[],\"articleType\":\"Method Article\",\"associatedPublications\":[],\"authors\":[{\"id\":437504821,\"identity\":\"b3b800f8-ecf9-4093-b267-59a4ddf9a92a\",\"order_by\":0,\"name\":\"Franz Chuquirachi\",\"email\":\"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAABEElEQVRIiWNgGAWjYBAC9gYInQAiDoAYbGAhAwucWngOYGgBCxlIEKcFwpAAs/FokT787OEPBrs8g+M5hocrauzy+CSfX93wo0CCgb+9OwGrFr40c6AjkosNzrwxOHjmWHIxm3RO2c0eoJjEmbMbsGmx52EwkzBgOJC44UaOwcEGtgOJbdI5aTd4gFoMJHKxauHhYf8GdDxMyz+gFskzaTf/4NXCYyZxAKalsQ2oRYL92G38tvCUSTYYJCfOPPOs4GBjX3JiG08O220ZAwkeXH4BOmyb5I8Ku8S+48mbPzZ8s0uc33782c03f2zk+Nt7sWqBAAMQkQA3Bszlwa0cDuBa2B8QoXoUjIJRMApGEAAAJVlkWD3lNosAAAAASUVORK5CYII=\",\"orcid\":\"\",\"institution\":\"Graduate School of System Design, Tokyo Metropolitan University\",\"correspondingAuthor\":true,\"prefix\":\"\",\"firstName\":\"Franz\",\"middleName\":\"\",\"lastName\":\"Chuquirachi\",\"suffix\":\"\"},{\"id\":437504822,\"identity\":\"9c944db0-06ec-4e20-9a95-5ed55d15a325\",\"order_by\":1,\"name\":\"Chyan Zheng Siow\",\"email\":\"\",\"orcid\":\"\",\"institution\":\"Graduate School of System Design, Tokyo Metropolitan University\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Chyan\",\"middleName\":\"Zheng\",\"lastName\":\"Siow\",\"suffix\":\"\"},{\"id\":437504823,\"identity\":\"19ebd5d8-1452-40c1-b58c-15fddb4ab345\",\"order_by\":2,\"name\":\"Wei Hong Chin\",\"email\":\"\",\"orcid\":\"\",\"institution\":\"Graduate School of System Design, Tokyo Metropolitan University\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Wei\",\"middleName\":\"Hong\",\"lastName\":\"Chin\",\"suffix\":\"\"},{\"id\":437504824,\"identity\":\"76cd7f0b-11f6-409e-9250-19438634a8ad\",\"order_by\":3,\"name\":\"Naoyuki Kubota\",\"email\":\"\",\"orcid\":\"\",\"institution\":\"Graduate School of System Design, Tokyo Metropolitan University\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Naoyuki\",\"middleName\":\"\",\"lastName\":\"Kubota\",\"suffix\":\"\"}],\"badges\":[],\"createdAt\":\"2025-04-02 14:36:30\",\"currentVersionCode\":1,\"declarations\":{\"humanSubjects\":true,\"vertebrateSubjects\":false,\"conflictsOfInterestStatement\":false,\"humanSubjectEthicalGuidelines\":true,\"humanSubjectConsent\":true,\"humanSubjectClinicalTrial\":false,\"humanSubjectCaseReport\":false,\"vertebrateSubjectEthicalGuidelines\":false},\"doi\":\"10.21203/rs.3.rs-6362240/v1\",\"doiUrl\":\"https://doi.org/10.21203/rs.3.rs-6362240/v1\",\"draftVersion\":[],\"editorialEvents\":[],\"editorialNote\":\"\",\"failedWorkflow\":false,\"files\":[{\"id\":79816180,\"identity\":\"67de06cd-f7e3-468e-9ff1-d912df886aa9\",\"added_by\":\"auto\",\"created_at\":\"2025-04-03 07:51:41\",\"extension\":\"pdf\",\"order_by\":1,\"title\":\"\",\"display\":\"\",\"copyAsset\":false,\"role\":\"manuscript-pdf\",\"size\":596364,\"visible\":true,\"origin\":\"\",\"legend\":\"\",\"description\":\"\",\"filename\":\"Paper2215.pdf\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-6362240/v1_covered_7d691d1e-2e7c-457a-a1e3-4a1cd5b25239.pdf\"}],\"financialInterests\":\"The authors declare no competing interests.\",\"formattedTitle\":\"\\u003cp\\u003eCosine Similarity of Convolutional Features of Video Frames as a Method for Finding Similar Activities from Past and Present Information of Video Input\\u003c/p\\u003e\",\"fulltext\":[],\"fulltextSource\":\"\",\"fullText\":\"\",\"funders\":[],\"hasAdminPriorityOnWorkflow\":false,\"hasManuscriptDocX\":false,\"hasOptedInToPreprint\":true,\"hasPassedJournalQc\":\"\",\"hasAnyPriority\":true,\"hideJournal\":true,\"highlight\":\"\",\"institution\":\"Tokyo Metropolitan University\",\"isAcceptedByJournal\":false,\"isAuthorSuppliedPdf\":true,\"isDeskRejected\":\"\",\"isHiddenFromSearch\":false,\"isInQc\":false,\"isInWorkflow\":false,\"isPdf\":true,\"isPdfUpToDate\":true,\"isWithdrawnOrRetracted\":false,\"journal\":{\"display\":true,\"email\":\"info@researchsquare.com\",\"identity\":\"researchsquare\",\"isNatureJournal\":false,\"hasQc\":true,\"allowDirectSubmit\":true,\"externalIdentity\":\"\",\"sideBox\":\"\",\"snPcode\":\"\",\"submissionUrl\":\"/submission\",\"title\":\"Research Square\",\"twitterHandle\":\"researchsquare\",\"acdcEnabled\":true,\"dfaEnabled\":false,\"editorialSystem\":\"\",\"reportingPortfolio\":\"\",\"inReviewEnabled\":false,\"inReviewRevisionsEnabled\":true},\"keywords\":\"Eldercare, Monitoring Systems, Activity Recognition, Human Activity Repetition, Cosine Similarity, Convolutional Neural Network, Video Analysis, Caregiver Assistance, Machine Learning\",\"lastPublishedDoi\":\"10.21203/rs.3.rs-6362240/v1\",\"lastPublishedDoiUrl\":\"https://doi.org/10.21203/rs.3.rs-6362240/v1\",\"license\":{\"name\":\"CC BY 4.0\",\"url\":\"https://creativecommons.org/licenses/by/4.0/\"},\"manuscriptAbstract\":\"\\u003cp\\u003eAmid the declining workforce in eldercare, researchers are expected to contribute with new technology and applications to improve productivity and reduce the complexity of the caregiver job. In the development of Monitoring Systems, the task of recognizing the need for help from images is vital. However, past solutions present some problems: solutions constrained to a list of cases, and dependency on complete recordings of activity. In this paper, we presented an original definition for a situation that requires help and proposed a method to solve the component related to finding repetitions of human activity, while overcoming the mentioned issues. We used the cosine similarity between convolutional features of frames of a video to find similar activity using only past and present information. We proved the effectiveness of the method by challenging it to count the number of repetitions of an activity in a video and comparing the results with the ground truth values. The resulting accuracy was 63.63%, showing signs that it can be used for constructing generalized solutions with more immediate output delivery.\\u003c/p\\u003e\",\"manuscriptTitle\":\"Cosine Similarity of Convolutional Features of Video Frames as a Method for Finding Similar Activities from Past and Present Information of Video Input\",\"msid\":\"\",\"msnumber\":\"\",\"nonDraftVersions\":[{\"code\":1,\"date\":\"2025-04-03 07:43:34\",\"doi\":\"10.21203/rs.3.rs-6362240/v1\",\"editorialEvents\":[{\"type\":\"communityComments\",\"content\":0}],\"status\":\"published\",\"journal\":{\"display\":true,\"email\":\"info@researchsquare.com\",\"identity\":\"researchsquare\",\"isNatureJournal\":false,\"hasQc\":true,\"allowDirectSubmit\":true,\"externalIdentity\":\"\",\"sideBox\":\"\",\"snPcode\":\"\",\"submissionUrl\":\"/submission\",\"title\":\"Research Square\",\"twitterHandle\":\"researchsquare\",\"acdcEnabled\":true,\"dfaEnabled\":false,\"editorialSystem\":\"\",\"reportingPortfolio\":\"\",\"inReviewEnabled\":false,\"inReviewRevisionsEnabled\":true}}],\"origin\":\"\",\"ownerIdentity\":\"0b9b3949-22e1-4c50-87ca-d9f2353e294e\",\"owner\":[],\"postedDate\":\"April 3rd, 2025\",\"published\":true,\"recentEditorialEvents\":[],\"rejectedJournal\":[],\"revision\":\"\",\"amendment\":\"\",\"status\":\"posted\",\"subjectAreas\":[{\"id\":46585879,\"name\":\"Artificial Intelligence and Machine Learning\"}],\"tags\":[],\"updatedAt\":\"2025-04-03T07:43:34+00:00\",\"versionOfRecord\":[],\"versionCreatedAt\":\"2025-04-03 07:43:34\",\"video\":\"\",\"vorDoi\":\"\",\"vorDoiUrl\":\"\",\"workflowStages\":[]},\"version\":\"v1\",\"identity\":\"rs-6362240\",\"journalConfig\":\"researchsquare\"},\"__N_SSP\":true},\"page\":\"/article/[identity]/[[...version]]\",\"query\":{\"redirect\":\"/article/rs-6362240\",\"identity\":\"rs-6362240\",\"version\":[\"v1\"]},\"buildId\":\"8U1c8b4HqxoKbykW_rLl7\",\"isFallback\":false,\"isExperimentalCompile\":false,\"dynamicIds\":[84888],\"gssp\":true,\"scriptLoader\":[]}","source_license":"CC-BY-4.0","license_restricted":false}