{"id":1078,"date":"2025-04-24T17:32:02","date_gmt":"2025-04-24T08:32:02","guid":{"rendered":"https:\/\/kmlab.nagaokaut.ac.jp\/?p=1078"},"modified":"2025-04-24T17:34:49","modified_gmt":"2025-04-24T08:34:49","slug":"our-paper-published-in-ieee-access","status":"publish","type":"post","link":"https:\/\/kmlab.nagaokaut.ac.jp\/index.php\/2025\/04\/24\/our-paper-published-in-ieee-access\/","title":{"rendered":"Our paper published in IEEE ACCESS"},"content":{"rendered":"

\"\"<\/h1>\n

Cross-Dataset Representation Learning for Unsupervised Deep Clustering in Human Activity Recognition<\/h1>\n

Authors: Tomoya Takatsu; Tessai Hayama; Hu Cui<\/h2>\n

https:\/\/ieeexplore.ieee.org\/document\/10971938<\/a><\/p>\n

Abstract:<\/h2>\n

This study presents a new method to improve unsupervised deep clustering for Human Activity Recognition (HAR). Traditional methods often struggle to extract meaningful features from unlabeled data, resulting in poor classification accuracy. To solve this, the proposed method combines an autoencoder with models pre-trained on diverse HAR datasets. This approach helps extract more reliable and generalizable feature representations, which are then used for clustering.<\/p>\n

The method was tested on three HAR datasets and significantly outperformed existing methods like k-means, HMM, and standard autoencoder-based clustering, achieving F1 scores between 0.441 and 0.781. Even with just 50 labeled samples for fine-tuning, the performance improved further, reaching F1 scores of 0.66 to 0.88. This demonstrates the method\u2019s robustness and effectiveness in unsupervised settings. It also shows promise for broader use beyond HAR.<\/p>\n

\u672c\u7814\u7a76\u306f\u3001\u4eba\u9593\u884c\u52d5\u8a8d\u8b58\uff08HAR\uff09\u306b\u304a\u3051\u308b\u6559\u5e2b\u306a\u3057\u6df1\u5c64\u30af\u30e9\u30b9\u30bf\u30ea\u30f3\u30b0\u306e\u7cbe\u5ea6\u3092\u5411\u4e0a\u3055\u305b\u308b\u65b0\u3057\u3044\u624b\u6cd5<\/strong>\u3092\u63d0\u6848\u3057\u3066\u3044\u307e\u3059\u3002\u5f93\u6765\u306e\u624b\u6cd5\u3067\u306f\u3001\u30e9\u30d9\u30eb\u306a\u3057\u30c7\u30fc\u30bf\u304b\u3089\u6709\u7528\u306a\u7279\u5fb4\u3092\u3046\u307e\u304f\u53d6\u308a\u51fa\u305b\u305a\u3001\u5206\u985e\u6027\u80fd\u304c\u4f4e\u3044\u3068\u3044\u3046\u8ab2\u984c\u304c\u3042\u308a\u307e\u3057\u305f\u3002<\/p>\n

\u305d\u3053\u3067\u672c\u7814\u7a76\u3067\u306f\u3001\u3055\u307e\u3056\u307e\u306aHAR\u30c7\u30fc\u30bf\u30bb\u30c3\u30c8\u3067\u4e8b\u524d\u5b66\u7fd2\u3055\u308c\u305f\u30e2\u30c7\u30eb\u3068\u30aa\u30fc\u30c8\u30a8\u30f3\u30b3\u30fc\u30c0\u3092\u7d44\u307f\u5408\u308f\u305b\u308b\u3053\u3068\u3067\u3001\u3088\u308a\u9811\u5065\u3067\u6c4e\u7528\u7684\u306a\u7279\u5fb4\u8868\u73fe\u3092\u62bd\u51fa<\/strong>\u3057\u3001\u305d\u308c\u3092\u4f7f\u3063\u3066\u30af\u30e9\u30b9\u30bf\u30ea\u30f3\u30b0\u3092\u884c\u3044\u307e\u3059\u3002<\/p>\n

3\u3064\u306eHAR\u30c7\u30fc\u30bf\u30bb\u30c3\u30c8\u3067\u691c\u8a3c\u3057\u305f\u7d50\u679c\u3001\u5f93\u6765\u624b\u6cd5\uff08k-means\u3084HMM\u3001\u901a\u5e38\u306e\u30aa\u30fc\u30c8\u30a8\u30f3\u30b3\u30fc\u30c0\uff09\u3088\u308a\u3082\u9ad8\u3044F1\u30b9\u30b3\u30a2\uff080.441\u301c0.781\uff09\u3092\u9054\u6210\u3002\u3055\u3089\u306b\u3001\u308f\u305a\u304b50\u500b\u306e\u30e9\u30d9\u30eb\u4ed8\u304d\u30c7\u30fc\u30bf\u3067\u5fae\u8abf\u6574\u3057\u305f\u5834\u5408\u3067\u3082F1\u30b9\u30b3\u30a2\u304c0.66\u301c0.88\u307e\u3067\u5411\u4e0a<\/strong>\u3057\u307e\u3057\u305f\u3002<\/p>\n

\u3053\u306e\u624b\u6cd5\u306f\u3001\u30e9\u30d9\u30eb\u306e\u5c11\u306a\u3044\u72b6\u6cc1\u3067\u3082\u9ad8\u3044\u8a8d\u8b58\u7cbe\u5ea6\u3092\u5b9f\u73fe<\/strong>\u3067\u304d\u308b\u3053\u3068\u3092\u793a\u3057\u3066\u304a\u308a\u3001HAR\u4ee5\u5916\u306e\u5206\u91ce\u306b\u3082\u5fdc\u7528\u53ef\u80fd\u306a\u3001\u5b9f\u7528\u6027\u3068\u62e1\u5f35\u6027\u306e\u9ad8\u3044\u65b9\u6cd5<\/strong>\u3067\u3059\u3002<\/p>\n\n\n

<\/p>\n","protected":false},"excerpt":{"rendered":"

Cross-Dataset Representation Learning for Unsupervised Deep Clustering in Human Activity Recognition Authors: … Continue reading Our paper published in IEEE ACCESS<\/span><\/a><\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"_exactmetrics_skip_tracking":false,"_exactmetrics_sitenote_active":false,"_exactmetrics_sitenote_note":"","_exactmetrics_sitenote_category":0,"_monsterinsights_skip_tracking":false,"_monsterinsights_sitenote_active":false,"_monsterinsights_sitenote_note":"","_monsterinsights_sitenote_category":0,"footnotes":""},"categories":[1],"tags":[],"class_list":["post-1078","post","type-post","status-publish","format-standard","hentry","category-1","without-featured-image"],"_links":{"self":[{"href":"https:\/\/kmlab.nagaokaut.ac.jp\/index.php\/wp-json\/wp\/v2\/posts\/1078","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/kmlab.nagaokaut.ac.jp\/index.php\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/kmlab.nagaokaut.ac.jp\/index.php\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/kmlab.nagaokaut.ac.jp\/index.php\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/kmlab.nagaokaut.ac.jp\/index.php\/wp-json\/wp\/v2\/comments?post=1078"}],"version-history":[{"count":3,"href":"https:\/\/kmlab.nagaokaut.ac.jp\/index.php\/wp-json\/wp\/v2\/posts\/1078\/revisions"}],"predecessor-version":[{"id":1082,"href":"https:\/\/kmlab.nagaokaut.ac.jp\/index.php\/wp-json\/wp\/v2\/posts\/1078\/revisions\/1082"}],"wp:attachment":[{"href":"https:\/\/kmlab.nagaokaut.ac.jp\/index.php\/wp-json\/wp\/v2\/media?parent=1078"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/kmlab.nagaokaut.ac.jp\/index.php\/wp-json\/wp\/v2\/categories?post=1078"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/kmlab.nagaokaut.ac.jp\/index.php\/wp-json\/wp\/v2\/tags?post=1078"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}