{"id":479752,"date":"2023-08-09T10:44:16","date_gmt":"2023-08-09T10:44:16","guid":{"rendered":""},"modified":"2023-09-05T11:19:30","modified_gmt":"2023-09-05T11:19:30","slug":"zero-shot-learning","status":"publish","type":"wiki","link":"https:\/\/oneproxy.pro\/jp\/wiki\/zero-shot-learning\/","title":{"rendered":"\u30bc\u30ed\u30b7\u30e7\u30c3\u30c8\u5b66\u7fd2"},"content":{"rendered":"<p>\u30bc\u30ed\u30b7\u30e7\u30c3\u30c8\u5b66\u7fd2\u306f\u3001\u4eba\u5de5\u77e5\u80fd\u3068\u6a5f\u68b0\u5b66\u7fd2\u306e\u5206\u91ce\u306b\u304a\u3051\u308b\u9769\u65b0\u7684\u306a\u6982\u5ff5\u3067\u3042\u308a\u3001\u30e2\u30c7\u30eb\u304c\u3053\u308c\u307e\u3067\u906d\u9047\u3057\u305f\u3053\u3068\u306e\u306a\u3044\u65b0\u3057\u3044\u30aa\u30d6\u30b8\u30a7\u30af\u30c8\u3084\u6982\u5ff5\u3092\u8a8d\u8b58\u3057\u3066\u7406\u89e3\u3067\u304d\u308b\u3088\u3046\u306b\u3057\u307e\u3059\u3002\u30e2\u30c7\u30eb\u304c\u81a8\u5927\u306a\u91cf\u306e\u30e9\u30d9\u30eb\u4ed8\u304d\u30c7\u30fc\u30bf\u3067\u30c8\u30ec\u30fc\u30cb\u30f3\u30b0\u3055\u308c\u308b\u5f93\u6765\u306e\u6a5f\u68b0\u5b66\u7fd2\u3068\u306f\u7570\u306a\u308a\u3001\u30bc\u30ed\u30b7\u30e7\u30c3\u30c8\u5b66\u7fd2\u3067\u306f\u3001\u660e\u793a\u7684\u306a\u30c8\u30ec\u30fc\u30cb\u30f3\u30b0\u306a\u3057\u3067\u3001\u65e2\u5b58\u306e\u77e5\u8b58\u304b\u3089\u65b0\u3057\u3044\u72b6\u6cc1\u306b\u6a5f\u68b0\u304c\u4e00\u822c\u5316\u3067\u304d\u307e\u3059\u3002<\/p>\n<h2>\u30bc\u30ed\u30b7\u30e7\u30c3\u30c8\u5b66\u7fd2\u306e\u8d77\u6e90\u3068\u305d\u306e\u6700\u521d\u306e\u8a00\u53ca\u306e\u6b74\u53f2<\/h2>\n<p>\u30bc\u30ed\u30b7\u30e7\u30c3\u30c8\u5b66\u7fd2\u306e\u8d77\u6e90\u306f\u3001\u7814\u7a76\u8005\u304c\u30bf\u30b9\u30af\u9593\u3067\u77e5\u8b58\u3092\u8ee2\u9001\u3059\u308b\u65b9\u6cd5\u3092\u6a21\u7d22\u3057\u59cb\u3081\u305f 2000 \u5e74\u4ee3\u521d\u982d\u306b\u9061\u308a\u307e\u3059\u30022009 \u5e74\u3001\u7814\u7a76\u8005\u306e Dolores Parra \u6c0f\u3068 Antonio Torralba \u6c0f\u306f\u3001\u8ad6\u6587\u300c\u610f\u5473\u8a18\u8ff0\u304b\u3089\u306e\u30bc\u30ed\u30b7\u30e7\u30c3\u30c8\u5b66\u7fd2\u300d\u3067\u300c\u30bc\u30ed\u30b7\u30e7\u30c3\u30c8\u5b66\u7fd2\u300d\u3068\u3044\u3046\u7528\u8a9e\u3092\u5c0e\u5165\u3057\u307e\u3057\u305f\u3002\u3053\u306e\u72ec\u5275\u7684\u306a\u7814\u7a76\u306f\u3001\u3053\u306e\u5206\u91ce\u306b\u304a\u3051\u308b\u305d\u306e\u5f8c\u306e\u9032\u6b69\u306e\u57fa\u790e\u3092\u7bc9\u304d\u307e\u3057\u305f\u3002<\/p>\n<h2>\u30bc\u30ed\u30b7\u30e7\u30c3\u30c8\u5b66\u7fd2\u306b\u95a2\u3059\u308b\u8a73\u7d30\u60c5\u5831\u3002\u30bc\u30ed\u30b7\u30e7\u30c3\u30c8\u5b66\u7fd2\u306e\u30c8\u30d4\u30c3\u30af\u3092\u62e1\u5f35\u3057\u307e\u3059\u3002<\/h2>\n<p>\u30bc\u30ed\u30b7\u30e7\u30c3\u30c8\u5b66\u7fd2\u306f\u3001\u5f93\u6765\u306e\u6a5f\u68b0\u5b66\u7fd2\u306e\u5927\u304d\u306a\u5236\u9650\u3067\u3042\u308b\u3001\u5e83\u7bc4\u306a\u30e9\u30d9\u30eb\u4ed8\u304d\u30c7\u30fc\u30bf\u306e\u5fc5\u8981\u6027\u306b\u5bfe\u51e6\u3057\u307e\u3059\u3002\u5f93\u6765\u306e\u6559\u5e2b\u3042\u308a\u5b66\u7fd2\u3067\u306f\u3001\u30e2\u30c7\u30eb\u306f\u906d\u9047\u3059\u308b\u53ef\u80fd\u6027\u306e\u3042\u308b\u3059\u3079\u3066\u306e\u30af\u30e9\u30b9\u306e\u4f8b\u3092\u5fc5\u8981\u3068\u3057\u307e\u3059\u3002\u4e00\u65b9\u3001\u30bc\u30ed\u30b7\u30e7\u30c3\u30c8\u5b66\u7fd2\u306f\u3001\u610f\u5473\u5c5e\u6027\u3001\u30c6\u30ad\u30b9\u30c8\u306e\u8aac\u660e\u3001\u307e\u305f\u306f\u95a2\u9023\u3059\u308b\u6982\u5ff5\u306a\u3069\u306e\u88dc\u52a9\u60c5\u5831\u3092\u6d3b\u7528\u3057\u3066\u3001\u65e2\u77e5\u3068\u672a\u77e5\u306e\u30ab\u30c6\u30b4\u30ea\u9593\u306e\u30ae\u30e3\u30c3\u30d7\u3092\u57cb\u3081\u307e\u3059\u3002<\/p>\n<h2>\u30bc\u30ed\u30b7\u30e7\u30c3\u30c8\u5b66\u7fd2\u306e\u5185\u90e8\u69cb\u9020\u3002\u30bc\u30ed\u30b7\u30e7\u30c3\u30c8\u5b66\u7fd2\u306e\u4ed5\u7d44\u307f\u3002<\/h2>\n<p>\u30bc\u30ed\u30b7\u30e7\u30c3\u30c8\u5b66\u7fd2\u306b\u306f\u8907\u6570\u306e\u30b9\u30c6\u30c3\u30d7\u304b\u3089\u306a\u308b\u30d7\u30ed\u30bb\u30b9\u304c\u542b\u307e\u308c\u307e\u3059\u3002<\/p>\n<ol>\n<li><strong>\u30bb\u30de\u30f3\u30c6\u30a3\u30c3\u30af\u57cb\u3081\u8fbc\u307f<\/strong>\u30c7\u30fc\u30bf \u30dd\u30a4\u30f3\u30c8\u3068\u30af\u30e9\u30b9\u306f\u3001\u305d\u308c\u3089\u306e\u95a2\u4fc2\u304c\u30ad\u30e3\u30d7\u30c1\u30e3\u3055\u308c\u308b\u5171\u901a\u306e\u30bb\u30de\u30f3\u30c6\u30a3\u30c3\u30af\u7a7a\u9593\u306b\u57cb\u3081\u8fbc\u307e\u308c\u307e\u3059\u3002<\/li>\n<li><strong>\u5c5e\u6027\u5b66\u7fd2<\/strong>\u30e2\u30c7\u30eb\u306f\u3001\u5404\u30af\u30e9\u30b9\u306b\u95a2\u9023\u4ed8\u3051\u3089\u308c\u305f\u30bb\u30de\u30f3\u30c6\u30a3\u30c3\u30af\u5c5e\u6027\u3092\u8a8d\u8b58\u3059\u308b\u3088\u3046\u306b\u30c8\u30ec\u30fc\u30cb\u30f3\u30b0\u3055\u308c\u307e\u3059\u3002<\/li>\n<li><strong>\u30bc\u30ed\u30b7\u30e7\u30c3\u30c8\u4e88\u6e2c<\/strong>\u65b0\u3057\u3044\u30af\u30e9\u30b9\u306b\u906d\u9047\u3059\u308b\u3068\u3001\u30e2\u30c7\u30eb\u306f\u5c5e\u6027\u30d9\u30fc\u30b9\u306e\u63a8\u8ad6\u3092\u4f7f\u7528\u3057\u3066\u3001\u4e8b\u524d\u306e\u30c8\u30ec\u30fc\u30cb\u30f3\u30b0 \u30c7\u30fc\u30bf\u304c\u306a\u304f\u3066\u3082\u305d\u306e\u30af\u30e9\u30b9\u306e\u7279\u5fb4\u3068\u5c5e\u6027\u3092\u4e88\u6e2c\u3057\u307e\u3059\u3002<\/li>\n<\/ol>\n<h2>\u30bc\u30ed\u30b7\u30e7\u30c3\u30c8\u5b66\u7fd2\u306e\u4e3b\u306a\u7279\u5fb4\u306e\u5206\u6790\u3002<\/h2>\n<p>\u30bc\u30ed\u30b7\u30e7\u30c3\u30c8\u5b66\u7fd2\u306e\u4e3b\u306a\u7279\u5fb4\u306f\u6b21\u306e\u3068\u304a\u308a\u3067\u3059\u3002<\/p>\n<ul>\n<li><strong>\u4e00\u822c\u5316<\/strong>\u30e2\u30c7\u30eb\u306f\u6700\u5c0f\u9650\u306e\u30c7\u30fc\u30bf\u3067\u65b0\u3057\u3044\u30af\u30e9\u30b9\u3092\u8a8d\u8b58\u3067\u304d\u308b\u305f\u3081\u3001\u8fc5\u901f\u306a\u9069\u5fdc\u304c\u53ef\u80fd\u306b\u306a\u308a\u307e\u3059\u3002<\/li>\n<li><strong>\u610f\u5473\u7406\u89e3<\/strong>: \u610f\u5473\u5c5e\u6027\u3068\u8aac\u660e\u3092\u4f7f\u7528\u3059\u308b\u3068\u3001\u5fae\u5999\u306a\u30cb\u30e5\u30a2\u30f3\u30b9\u3092\u7406\u89e3\u3057\u3084\u3059\u304f\u306a\u308a\u307e\u3059\u3002<\/li>\n<li><strong>\u30c7\u30fc\u30bf\u4f9d\u5b58\u6027\u306e\u4f4e\u6e1b<\/strong>\u30bc\u30ed\u30b7\u30e7\u30c3\u30c8\u5b66\u7fd2\u306b\u3088\u308a\u3001\u5e83\u7bc4\u306a\u30e9\u30d9\u30eb\u4ed8\u304d\u30c7\u30fc\u30bf\u306e\u5fc5\u8981\u6027\u304c\u6e1b\u308a\u3001\u30c7\u30fc\u30bf\u53d6\u5f97\u30b3\u30b9\u30c8\u304c\u524a\u6e1b\u3055\u308c\u307e\u3059\u3002<\/li>\n<\/ul>\n<h2>\u30bc\u30ed\u30b7\u30e7\u30c3\u30c8\u5b66\u7fd2\u306e\u7a2e\u985e<\/h2>\n<p>\u30bc\u30ed\u30b7\u30e7\u30c3\u30c8\u5b66\u7fd2\u30a2\u30d7\u30ed\u30fc\u30c1\u306b\u306f\u3044\u304f\u3064\u304b\u306e\u7a2e\u985e\u304c\u3042\u308a\u307e\u3059\u3002<\/p>\n<ol>\n<li><strong>\u5c5e\u6027\u30d9\u30fc\u30b9<\/strong>\u30e2\u30c7\u30eb\u306f\u30af\u30e9\u30b9\u306b\u95a2\u9023\u4ed8\u3051\u3089\u308c\u305f\u5c5e\u6027\u3092\u4e88\u6e2c\u3057\u3001\u305d\u308c\u3092\u4f7f\u7528\u3057\u3066\u6a5f\u80fd\u3092\u63a8\u6e2c\u3057\u307e\u3059\u3002<\/li>\n<li><strong>\u30bb\u30de\u30f3\u30c6\u30a3\u30c3\u30af\u30d9\u30fc\u30b9<\/strong>\u30af\u30e9\u30b9\u3068\u30a4\u30f3\u30b9\u30bf\u30f3\u30b9\u9593\u306e\u610f\u5473\u7684\u95a2\u4fc2\u3092\u6d3b\u7528\u3057\u3066\u4e88\u6e2c\u3092\u884c\u3044\u307e\u3059\u3002<\/li>\n<li><strong>\u30cf\u30a4\u30d6\u30ea\u30c3\u30c9\u30a2\u30d7\u30ed\u30fc\u30c1<\/strong>: \u8907\u6570\u306e\u88dc\u52a9\u60c5\u5831\u30bd\u30fc\u30b9\u3092\u7d44\u307f\u5408\u308f\u305b\u3066\u3001\u3088\u308a\u6b63\u78ba\u306a\u4e88\u6e2c\u3092\u884c\u3044\u307e\u3059\u3002<\/li>\n<\/ol>\n<p>\u305d\u308c\u305e\u308c\u306e\u7279\u5fb4\u3092\u307e\u3068\u3081\u305f\u8868\u3092\u4ee5\u4e0b\u306b\u793a\u3057\u307e\u3059\u3002<\/p>\n<table>\n<thead>\n<tr>\n<th>\u30a2\u30d7\u30ed\u30fc\u30c1<\/th>\n<th>\u8aac\u660e<\/th>\n<\/tr>\n<\/thead>\n<tbody>\n<tr>\n<td>\u5c5e\u6027\u30d9\u30fc\u30b9<\/td>\n<td>\u30af\u30e9\u30b9\u306e\u5c5e\u6027\u306e\u4e88\u6e2c\u306b\u7126\u70b9\u3092\u5f53\u3066\u307e\u3059\u3002<\/td>\n<\/tr>\n<tr>\n<td>\u30bb\u30de\u30f3\u30c6\u30a3\u30c3\u30af\u30d9\u30fc\u30b9<\/td>\n<td>\u63a8\u8ad6\u306b\u610f\u5473\u95a2\u4fc2\u3092\u5229\u7528\u3057\u307e\u3059\u3002<\/td>\n<\/tr>\n<tr>\n<td>\u30cf\u30a4\u30d6\u30ea\u30c3\u30c9\u30a2\u30d7\u30ed\u30fc\u30c1<\/td>\n<td>\u8907\u6570\u306e\u30bd\u30fc\u30b9\u3092\u7d44\u307f\u5408\u308f\u305b\u3066\u7cbe\u5ea6\u3092\u9ad8\u3081\u307e\u3059\u3002<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<h2>Zero-shot Learning \u306e\u4f7f\u3044\u65b9\u3001\u4f7f\u7528\u4e0a\u306e\u554f\u984c\u70b9\u3068\u305d\u306e\u89e3\u6c7a\u7b56\u3002<\/h2>\n<p>\u30bc\u30ed\u30b7\u30e7\u30c3\u30c8\u5b66\u7fd2\u306f\u3055\u307e\u3056\u307e\u306a\u5206\u91ce\u3067\u5fdc\u7528\u3055\u308c\u3066\u3044\u307e\u3059\u3002<\/p>\n<ul>\n<li><strong>\u753b\u50cf\u8a8d\u8b58<\/strong>: \u753b\u50cf\u5185\u306e\u65b0\u3057\u3044\u30aa\u30d6\u30b8\u30a7\u30af\u30c8\u3092\u8b58\u5225\u3057\u307e\u3059\u3002<\/li>\n<li><strong>\u81ea\u7136\u8a00\u8a9e\u51e6\u7406<\/strong>: \u672a\u77e5\u306e\u30c8\u30d4\u30c3\u30af\u306b\u95a2\u3059\u308b\u30c6\u30ad\u30b9\u30c8\u3092\u7406\u89e3\u3057\u3001\u751f\u6210\u3057\u307e\u3059\u3002<\/li>\n<li><strong>\u533b\u7642\u753b\u50cf\u51e6\u7406<\/strong>: \u65b0\u3057\u3044\u75c5\u6c17\u306e\u72b6\u614b\u3092\u8a3a\u65ad\u3057\u307e\u3059\u3002<\/li>\n<\/ul>\n<p>\u8ab2\u984c\u3068\u3057\u3066\u306f\u3001\u30c7\u30fc\u30bf\u306e\u5e0c\u8584\u6027\u3068\u7cbe\u5ea6\u306e\u5236\u9650\u304c\u6319\u3052\u3089\u308c\u307e\u3059\u3002\u89e3\u6c7a\u7b56\u3068\u3057\u3066\u306f\u3001\u5c5e\u6027\u6ce8\u91c8\u306e\u6539\u5584\u3068\u30bb\u30de\u30f3\u30c6\u30a3\u30c3\u30af\u57cb\u3081\u8fbc\u307f\u306e\u6539\u5584\u304c\u6319\u3052\u3089\u308c\u307e\u3059\u3002<\/p>\n<h2>\u4e3b\u306a\u7279\u5fb4\u3068\u985e\u4f3c\u7528\u8a9e\u3068\u306e\u6bd4\u8f03<\/h2>\n<table>\n<thead>\n<tr>\n<th>\u7279\u6027<\/th>\n<th>\u30bc\u30ed\u30b7\u30e7\u30c3\u30c8\u5b66\u7fd2<\/th>\n<th>\u8ee2\u79fb\u5b66\u7fd2<\/th>\n<th>\u5c11\u6570\u30b7\u30e7\u30c3\u30c8\u5b66\u7fd2<\/th>\n<\/tr>\n<\/thead>\n<tbody>\n<tr>\n<td>\u65b0\u3057\u3044\u30bf\u30b9\u30af\u3078\u306e\u9069\u5fdc\u529b<\/td>\n<td>\u9ad8\u3044<\/td>\n<td>\u9069\u5ea6<\/td>\n<td>\u9069\u5ea6<\/td>\n<\/tr>\n<tr>\n<td>\u30e9\u30d9\u30eb\u4ed8\u304d\u30c7\u30fc\u30bf\u306e\u8981\u4ef6<\/td>\n<td>\u4f4e\u3044<\/td>\n<td>\u4e2d\u7a0b\u5ea6\u304b\u3089\u9ad8\u7a0b\u5ea6<\/td>\n<td>\u4f4e\u3044<\/td>\n<\/tr>\n<tr>\n<td>\u4e00\u822c\u5316\u80fd\u529b<\/td>\n<td>\u9ad8\u3044<\/td>\n<td>\u9ad8\u3044<\/td>\n<td>\u9069\u5ea6<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<h2>\u30bc\u30ed\u30b7\u30e7\u30c3\u30c8\u5b66\u7fd2\u306b\u95a2\u3059\u308b\u5c06\u6765\u306e\u5c55\u671b\u3068\u6280\u8853\u3002<\/h2>\n<p>\u30bc\u30ed\u30b7\u30e7\u30c3\u30c8\u5b66\u7fd2\u306e\u5c06\u6765\u306b\u306f\u3001\u523a\u6fc0\u7684\u306a\u53ef\u80fd\u6027\u304c\u79d8\u3081\u3089\u308c\u3066\u3044\u307e\u3059\u3002<\/p>\n<ul>\n<li><strong>\u30e1\u30bf\u5b66\u7fd2<\/strong>: \u5b66\u7fd2\u65b9\u6cd5\u3092\u5b66\u7fd2\u3057\u3001\u9069\u5fdc\u3092\u52a0\u901f\u3059\u308b\u30e2\u30c7\u30eb\u3002<\/li>\n<li><strong>\u30bc\u30ed\u30b7\u30e7\u30c3\u30c8\u5f37\u5316\u5b66\u7fd2<\/strong>: \u5f37\u5316\u5b66\u7fd2\u3068\u30bc\u30ed\u30b7\u30e7\u30c3\u30c8\u30d1\u30e9\u30c0\u30a4\u30e0\u3092\u878d\u5408\u3057\u307e\u3059\u3002<\/li>\n<li><strong>\u30bc\u30ed\u30b7\u30e7\u30c3\u30c8\u30de\u30eb\u30c1\u30e2\u30fc\u30c0\u30eb\u30d5\u30e5\u30fc\u30b8\u30e7\u30f3<\/strong>: \u30bc\u30ed\u30b7\u30e7\u30c3\u30c8\u5b66\u7fd2\u3092\u8907\u6570\u306e\u30c7\u30fc\u30bf\u30e2\u30c0\u30ea\u30c6\u30a3\u306b\u62e1\u5f35\u3057\u307e\u3059\u3002<\/li>\n<\/ul>\n<h2>\u30d7\u30ed\u30ad\u30b7 \u30b5\u30fc\u30d0\u30fc\u3092 Zero-shot Learning \u3067\u4f7f\u7528\u3059\u308b\u65b9\u6cd5\u3001\u307e\u305f\u306f Zero-shot Learning \u306b\u95a2\u9023\u4ed8\u3051\u308b\u65b9\u6cd5\u3002<\/h2>\n<p>\u30d7\u30ed\u30ad\u30b7 \u30b5\u30fc\u30d0\u30fc\u306f\u3001\u30bc\u30ed\u30b7\u30e7\u30c3\u30c8\u5b66\u7fd2\u30a2\u30d7\u30ea\u30b1\u30fc\u30b7\u30e7\u30f3\u3092\u5b9f\u73fe\u3059\u308b\u4e0a\u3067\u91cd\u8981\u306a\u5f79\u5272\u3092\u679c\u305f\u3057\u307e\u3059\u3002<\/p>\n<ul>\n<li><strong>\u30c7\u30fc\u30bf\u53ce\u96c6<\/strong>: \u30d7\u30ed\u30ad\u30b7 \u30b5\u30fc\u30d0\u30fc\u3092\u4f7f\u7528\u3059\u308b\u3068\u3001\u3055\u307e\u3056\u307e\u306a\u5730\u7406\u7684\u5730\u57df\u304b\u3089\u591a\u69d8\u306a\u30c7\u30fc\u30bf\u3092\u53ce\u96c6\u3057\u3001\u30c8\u30ec\u30fc\u30cb\u30f3\u30b0 \u30d7\u30ed\u30bb\u30b9\u3092\u5f37\u5316\u3067\u304d\u307e\u3059\u3002<\/li>\n<li><strong>\u30d7\u30e9\u30a4\u30d0\u30b7\u30fc\u4fdd\u8b77<\/strong>: \u30d7\u30ed\u30ad\u30b7 \u30b5\u30fc\u30d0\u30fc\u306f\u3001\u30c7\u30fc\u30bf\u8981\u6c42\u306e\u9001\u4fe1\u5143\u3092\u30de\u30b9\u30af\u3059\u308b\u3053\u3068\u3067\u30c7\u30fc\u30bf\u306e\u30d7\u30e9\u30a4\u30d0\u30b7\u30fc\u3092\u5f37\u5316\u3057\u3001\u30c7\u30fc\u30bf\u4fdd\u8b77\u898f\u5236\u3078\u306e\u6e96\u62e0\u3092\u4fdd\u8a3c\u3057\u307e\u3059\u3002<\/li>\n<\/ul>\n<h2>\u95a2\u9023\u30ea\u30f3\u30af<\/h2>\n<p>\u30bc\u30ed\u30b7\u30e7\u30c3\u30c8\u5b66\u7fd2\u306e\u8a73\u7d30\u306b\u3064\u3044\u3066\u306f\u3001\u6b21\u306e\u30ea\u30bd\u30fc\u30b9\u3092\u53c2\u7167\u3057\u3066\u304f\u3060\u3055\u3044\u3002<\/p>\n<ul>\n<li><a href=\"\/jp\/link-to-paper\/\" target=\"_new\" rel=\"noopener\">\u30c9\u30ed\u30ec\u30b9\u30fb\u30d1\u30e9\u3068\u30a2\u30f3\u30c8\u30cb\u30aa\u30fb\u30c8\u30c3\u30e9\u30eb\u30d0\u306e\u539f\u8457\u8ad6\u6587<\/a><\/li>\n<li><a href=\"\/jp\/link-to-survey\/\" target=\"_new\" rel=\"noopener\">\u30bc\u30ed\u30b7\u30e7\u30c3\u30c8\u5b66\u7fd2\uff1a\u5305\u62ec\u7684\u306a\u8abf\u67fb<\/a><\/li>\n<li><a href=\"\/jp\/link-to-advances\/\" target=\"_new\" rel=\"noopener\">\u30bc\u30ed\u30b7\u30e7\u30c3\u30c8\u5b66\u7fd2\u6280\u8853\u306e\u9032\u6b69<\/a><\/li>\n<\/ul>\n<p>\u6a5f\u68b0\u5b66\u7fd2\u306e\u5206\u91ce\u304c\u9032\u5316\u3092\u7d9a\u3051\u308b\u4e2d\u3001\u30bc\u30ed\u30b7\u30e7\u30c3\u30c8\u5b66\u7fd2\u306f\u57fa\u790e\u3068\u3057\u3066\u969b\u7acb\u3063\u3066\u304a\u308a\u3001\u304b\u3064\u3066\u306f\u4e0d\u53ef\u80fd\u3068\u601d\u308f\u308c\u3066\u3044\u305f\u65b9\u6cd5\u3067\u6a5f\u68b0\u304c\u5b66\u7fd2\u3057\u3001\u9069\u5fdc\u3059\u308b\u3053\u3068\u3092\u53ef\u80fd\u306b\u3057\u307e\u3059\u3002\u30d7\u30ed\u30ad\u30b7 \u30b5\u30fc\u30d0\u30fc\u306a\u3069\u306e\u30c6\u30af\u30ce\u30ed\u30b8\u306e\u30b5\u30dd\u30fc\u30c8\u306b\u3088\u308a\u3001\u771f\u306b\u30a4\u30f3\u30c6\u30ea\u30b8\u30a7\u30f3\u30c8\u306a\u30b7\u30b9\u30c6\u30e0\u3078\u306e\u9053\u306f\u3001\u3053\u308c\u307e\u3067\u4ee5\u4e0a\u306b\u5b9f\u73fe\u53ef\u80fd\u306b\u306a\u308a\u307e\u3059\u3002<\/p>","protected":false},"featured_media":470992,"menu_order":0,"template":"","meta":{"_acf_changed":false,"content-type":"","inline_featured_image":false,"footnotes":""},"class_list":["post-479752","wiki","type-wiki","status-publish","has-post-thumbnail","hentry"],"acf":{"faq_title":"Frequently Asked Questions about <mark>Zero-shot Learning: Bridging the Gap between Knowledge and Adaptability<\/mark>","faq_items":[{"question":"What is Zero-shot Learning?","answer":"Zero-shot learning is a revolutionary approach in artificial intelligence and machine learning. Unlike traditional methods that require extensive labeled data for each new class, zero-shot learning allows models to generalize and recognize new concepts they haven't been directly trained on. This is achieved by leveraging auxiliary information like semantic attributes and descriptions."},{"question":"How did Zero-shot Learning originate?","answer":"The concept of Zero-shot Learning dates back to the early 2000s. In 2009, researchers Dolores Parra and Antonio Torralba coined the term in their paper \"Zero-Shot Learning from Semantic Descriptions.\" This marked the beginning of exploring ways to enable models to adapt and learn from novel classes without explicit training."},{"question":"How does Zero-shot Learning work?","answer":"Zero-shot learning involves several steps:\r\n<ol>\r\n \t<li><strong>Semantic Embeddings<\/strong>: Data and classes are embedded in a semantic space.<\/li>\r\n \t<li><strong>Attribute Learning<\/strong>: Models learn to predict attributes of classes.<\/li>\r\n \t<li><strong>Zero-shot Prediction<\/strong>: When encountering a new class, the model uses attributes to predict features.<\/li>\r\n<\/ol>"},{"question":"What are the key features of Zero-shot Learning?","answer":"Key features include:\r\n<ul>\r\n \t<li><strong>Generalization<\/strong>: Models can recognize new classes quickly.<\/li>\r\n \t<li><strong>Semantic Understanding<\/strong>: Using semantic attributes enhances nuanced comprehension.<\/li>\r\n \t<li><strong>Reduced Data Dependency<\/strong>: Less labeled data is needed, reducing data acquisition costs.<\/li>\r\n<\/ul>"},{"question":"What types of Zero-shot Learning exist?","answer":"There are several types:\r\n<ol>\r\n \t<li><strong>Attribute-based<\/strong>: Predicts attributes for class inference.<\/li>\r\n \t<li><strong>Semantic-based<\/strong>: Relies on semantic relationships.<\/li>\r\n \t<li><strong>Hybrid Approaches<\/strong>: Combines multiple sources of information.<\/li>\r\n<\/ol>"},{"question":"Where can Zero-shot Learning be applied?","answer":"Zero-shot learning finds applications in:\r\n<ul>\r\n \t<li><strong>Image Recognition<\/strong>: Identifying new objects in images.<\/li>\r\n \t<li><strong>Natural Language Processing<\/strong>: Understanding and generating text on unseen topics.<\/li>\r\n \t<li><strong>Medical Imaging<\/strong>: Diagnosing conditions for new diseases.<\/li>\r\n<\/ul>"},{"question":"What challenges does Zero-shot Learning face?","answer":"Challenges include data sparsity and accuracy limitations. Solutions involve better attribute annotation and improved semantic embeddings."},{"question":"How does Zero-shot Learning compare to Transfer Learning and Few-shot Learning?","answer":"<table>\r\n<thead>\r\n<tr>\r\n<th>Characteristic<\/th>\r\n<th>Zero-shot Learning<\/th>\r\n<th>Transfer Learning<\/th>\r\n<th>Few-shot Learning<\/th>\r\n<\/tr>\r\n<\/thead>\r\n<tbody>\r\n<tr>\r\n<td>Adaptability to New Tasks<\/td>\r\n<td>High<\/td>\r\n<td>Moderate<\/td>\r\n<td>Moderate<\/td>\r\n<\/tr>\r\n<tr>\r\n<td>Labeled Data Requirement<\/td>\r\n<td>Low<\/td>\r\n<td>Moderate to High<\/td>\r\n<td>Low<\/td>\r\n<\/tr>\r\n<tr>\r\n<td>Generalization Ability<\/td>\r\n<td>High<\/td>\r\n<td>High<\/td>\r\n<td>Moderate<\/td>\r\n<\/tr>\r\n<\/tbody>\r\n<\/table>"},{"question":"What does the future hold for Zero-shot Learning?","answer":"The future brings exciting prospects:\r\n<ul>\r\n \t<li><strong>Meta-learning<\/strong>: Models learn how to learn, speeding up adaptation.<\/li>\r\n \t<li><strong>Zero-shot Reinforcement Learning<\/strong>: Merging reinforcement learning with zero-shot paradigms.<\/li>\r\n \t<li><strong>Zero-shot Multimodal Fusion<\/strong>: Extending zero-shot learning across different data types.<\/li>\r\n<\/ul>"},{"question":"How are proxy servers related to Zero-shot Learning?","answer":"Proxy servers play a vital role:\r\n<ul>\r\n \t<li><strong>Data Collection<\/strong>: They gather diverse data from various regions, enriching training.<\/li>\r\n \t<li><strong>Privacy Protection<\/strong>: Proxy servers ensure data privacy by masking data request origins.<\/li>\r\n<\/ul>"}]},"_links":{"self":[{"href":"https:\/\/oneproxy.pro\/jp\/wp-json\/wp\/v2\/wiki\/479752","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/oneproxy.pro\/jp\/wp-json\/wp\/v2\/wiki"}],"about":[{"href":"https:\/\/oneproxy.pro\/jp\/wp-json\/wp\/v2\/types\/wiki"}],"version-history":[{"count":0,"href":"https:\/\/oneproxy.pro\/jp\/wp-json\/wp\/v2\/wiki\/479752\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/oneproxy.pro\/jp\/wp-json\/wp\/v2\/media\/470992"}],"wp:attachment":[{"href":"https:\/\/oneproxy.pro\/jp\/wp-json\/wp\/v2\/media?parent=479752"}],"curies":[{"name":"\u3046\u30fc\u3093","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}