{"id":3012,"date":"2023-03-06T05:37:00","date_gmt":"2023-03-05T20:37:00","guid":{"rendered":"https:\/\/devneko.jp\/wordpress\/?p=3012"},"modified":"2023-03-06T05:37:00","modified_gmt":"2023-03-05T20:37:00","slug":"language-is-not-all-you-need","status":"publish","type":"post","link":"https:\/\/devneko.jp\/wordpress\/?p=3012","title":{"rendered":"Language Is Not All You Need"},"content":{"rendered":"\n<ul class=\"wp-block-list\">\n<li><strong>Language Is Not All You Need: Aligning Perception with Language Models\u00a0<\/strong>[110.5]<br>Kosmos-1\u306fMLLM(Multimodal Large Language Model)\u3067\u3001\u4e00\u822c\u7684\u306a\u30e2\u30c0\u30ea\u30c6\u30a3\u3092\u8a8d\u8b58\u3057\u3001\u6587\u8108\u3067\u5b66\u7fd2\u3057\u3001\u6307\u793a\u306b\u5f93\u3046\u3053\u3068\u304c\u3067\u304d\u308b\u3002 \u6211\u3005\u306f\u3001\u4efb\u610f\u306b\u30a4\u30f3\u30bf\u30fc\u30ea\u30fc\u30d6\u3055\u308c\u305f\u30c6\u30ad\u30b9\u30c8\u3068\u753b\u50cf\u3001\u753b\u50cf\u30ad\u30e3\u30d7\u30c1\u30e3\u30da\u30a2\u3001\u30c6\u30ad\u30b9\u30c8\u30c7\u30fc\u30bf\u3092\u542b\u3080\u3001Web\u30b9\u30b1\u30fc\u30eb\u306e\u30de\u30eb\u30c1\u30e2\u30fc\u30c0\u30eb\u30b3\u30fc\u30d1\u30b9\u4e0a\u3067\u3001Kosmos-1\u3092\u30b9\u30af\u30e9\u30c3\u30c1\u304b\u3089\u30c8\u30ec\u30fc\u30cb\u30f3\u30b0\u3059\u308b\u3002 \u5b9f\u9a13\u7d50\u679c\u304b\u3089,Kosmos-1 \u306f (i) \u8a00\u8a9e\u7406\u89e3,\u751f\u6210,\u3055\u3089\u306b\u306f OCR \u30d5\u30ea\u30fc NLP \u306b\u304a\u3044\u3066\u512a\u308c\u305f\u6027\u80fd\u3092\u767a\u63ee\u3059\u308b\u3053\u3068\u304c\u793a\u3055\u308c\u305f\u3002 \u307e\u305f\u3001MLLM\u306f\u8a00\u8a9e\u304b\u3089\u30de\u30eb\u30c1\u30e2\u30fc\u30c0\u30eb\u3078\u306e\u77e5\u8b58\u306e\u4f1d\u9054\u3084\u591a\u30e2\u30fc\u30c0\u30eb\u304b\u3089\u8a00\u8a9e\u3078\u306e\u77e5\u8b58\u306e\u4f1d\u9054\u3068\u3044\u3063\u305f\u3001\u30af\u30ed\u30b9\u30e2\u30fc\u30c0\u30eb\u8ee2\u9001\u306e\u6069\u6075\u3092\u53d7\u3051\u308b\u3053\u3068\u304c\u3067\u304d\u308b\u3053\u3068\u3092\u793a\u3059\u3002<br><a href=\"http:\/\/arxiv.org\/abs\/2302.14045v2\">\u8ad6\u6587<\/a>\u00a0\u00a0<a href=\"https:\/\/fugumt.com\/fugumt\/paper_check\/2302.14045v2\">\u53c2\u8003\u8a33\uff08\u30e1\u30bf\u30c7\u30fc\u30bf\uff09<\/a>\u00a0 \u00a0(Wed, 1 Mar 2023 11:04:51 GMT)<\/li>\n\n\n\n<li>\u8a00\u8a9e\u30e2\u30c7\u30eb\u306e\u9650\u754c\uff08\u30c6\u30ad\u30b9\u30c8\uff09\u3092\u8d85\u3048\u308b\u30de\u30eb\u30c1\u30e2\u30fc\u30c0\u30eb\u30e2\u30c7\u30ebKOSMOS-1\u306e\u63d0\u6848\u3002Image Captioning\u3084VQA\u3067\u512a\u308c\u305f\u6027\u80fd\uff08Flamingo\u3068\u7af6\u5408\uff09\u3002Web-based Structural Reading Comprehension\u3067\u30de\u30eb\u30c1\u30e2\u30fc\u30c0\u30eb\u306e\u6069\u6075\u3092\u53d7\u3051\u308b\u306e\u306f\u78ba\u304b\u306b\u3001\u3068\u3044\u3046\u611f\u60f3\u3067\u5b9f\u7528\u6027\u3082\u3042\u308a\u305d\u3046\u3002<\/li>\n\n\n\n<li>\u00a0KOSMOS-1 is about 1.6B\u3068\u306e\u3053\u3068\u3067\u6628\u4eca\u306e\u5927\u898f\u6a21\u8a00\u8a9e\u30e2\u30c7\u30eb\u306b\u6bd4\u3079\u308b\u3068\u975e\u5e38\u306b\u5927\u304d\u3044\u3068\u3082\u8a00\u3048\u306a\u3044\u30b5\u30a4\u30ba\u3002<\/li>\n\n\n\n<li>\u8ad6\u6587\u306b\u306f<a href=\"https:\/\/github.com\/microsoft\/unilm\">GitHub &#8211; microsoft\/unilm: Large-scale Self-supervised Pre-training Across Tasks, Languages, and Modalities<\/a>\u304c\u66f8\u304b\u308c\u3066\u3044\u308b\u304c\u73fe\u6642\u70b9\u3067\u306f\u516c\u958b\u8e0f\u7834\u3055\u308c\u3066\u3044\u306a\u3044\u3088\u3046<\/li>\n<\/ul>\n","protected":false},"excerpt":{"rendered":"","protected":false},"author":1,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[2],"tags":[313,524],"class_list":["post-3012","post","type-post","status-publish","format-standard","hentry","category-arxiv","tag-prompt","tag-524"],"_links":{"self":[{"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=\/wp\/v2\/posts\/3012","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=3012"}],"version-history":[{"count":0,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=\/wp\/v2\/posts\/3012\/revisions"}],"wp:attachment":[{"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=3012"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=3012"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=3012"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}