{"id":10941,"date":"2026-08-11T18:28:05","date_gmt":"2026-08-11T10:28:05","guid":{"rendered":"https:\/\/infernews.com\/blog\/concor\/"},"modified":"2026-08-11T18:31:55","modified_gmt":"2026-08-11T10:31:55","slug":"concor","status":"publish","type":"post","link":"https:\/\/infernews.com\/blog\/concor\/","title":{"rendered":"ConCor-1\uff1a\u4e00\u53e5\u6a19\u8a3b\u90fd\u5514\u4f7f\u7540\uff0c\u81ea\u52d5\u6435\u51fa\u5716\u6587\u5c0d\u61c9"},"content":{"rendered":"\n<figure class=\"wp-block-image size-large\"><img decoding=\"async\" src=\"https:\/\/infernews.com\/blog\/wp-content\/uploads\/2026\/08\/pasted-fad15cd9baa0.jpg\" alt=\"Bidirectional concept correspondence: a caption, a referring expression and a category list all produce the same output\"><\/figure>\n\n\n\n<p class=\"wp-block-paragraph\">\u4ee5\u5f80\u505a\u8996\u89ba\u8a9e\u8a00\u5b9a\u4f4d\uff0c\u591a\u6578\u6d41\u7a0b\u90fd\u8981\u4f60\u5148\u8b1b\u660e\u60f3\u6435\u908a\u53e5\u5b57\uff0c\u6a21\u578b\u518d\u55ba\u5716\u5165\u9762\u6307\u51fa\u5c0d\u61c9\u5340\u57df\u3002ConCor-1 \u7d22\u6027\u53cd\u8f49\u5462\u500b\u505a\u6cd5\uff1a\u7540\u4e00\u5f35\u5716\u540c\u4e00\u6bb5\u6587\u5b57\uff0c\u7121\u8ad6\u4fc2\u5b8c\u6574\u63cf\u8ff0\u3001\u6307\u4ee3\u8868\u9054\uff0c\u5b9a\u4fc2\u4e00\u5217\u985e\u5225\u540d\uff0c\u6a21\u578b\u90fd\u6703\u81ea\u5df1\u5224\u65b7\u908a\u6bb5\u6587\u5b57\u540c\u5716\u5165\u9762\u908a\u500b\u7269\u4ef6\u5c0d\u5f97\u4e0a\uff0c\u7136\u5f8c\u4e00\u6b21\u904e\u7540\u9f4a\u6587\u5b57\u906e\u7f69\u3001\u5be6\u4f8b\u906e\u7f69\u540c\u5c0d\u61c9\u5206\u6578\u3002\u7814\u7a76\u5718\u968a\u7531\u83ef\u76db\u9813\u5927\u5b78\u3001Allen Institute for AI \u540c Meta FAIR \u7d44\u6210\uff0c\u6a21\u578b\u57fa\u65bc\u9810\u8a13\u7df4\u8996\u89ba\u8a9e\u8a00\u6a21\u578b\uff0c\u518d\u52a0\u4e0a\u4e00\u7d44\u53ef\u5b78\u7fd2\u5605 bridge tokens \u569f\u4ee3\u8868\u5019\u9078\u5c0d\u61c9\u3002<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">\u5462\u7a2e\u8a2d\u8a08\u6700\u5927\u5605\u5be6\u7528\u50f9\u503c\uff0c\u4fc2\u7701\u537b\u524d\u7f6e\u6a19\u8a3b\u5605\u5de5\u5e8f\u3002\u7576\u4f60\u624b\u4e0a\u5f97\u4e00\u6bb5\u9577 caption \u6216\u8005\u4e00\u5927\u4e32\u985e\u5225\u540d\uff0c\u50b3\u7d71\u65b9\u6cd5\u5f80\u5f80\u8981\u9010\u53e5\u62c6\u958b\u518d\u5206\u6279\u9935\u5165\uff0cConCor-1 \u76f4\u63a5\u7576\u4f5c\u4e00\u6b21\u5c0d\u61c9\u9810\u6e2c\u8655\u7406\uff0c\u6587\u5b57\u5206\u5272\u3001\u5f71\u50cf\u5206\u5272\u3001\u8de8\u6a21\u614b\u5c0d\u9f4a\u4e09\u4ef6\u4e8b\u4e00\u9f4a\u505a\u3002\u4f62\u5c07 phrase grounding\u3001referring expression \u540c open-vocabulary detection \u6536\u7d0d\u6210\u540c\u4e00\u683c\u5f0f\uff0c\u7b49\u8a13\u7df4\u540c\u8a55\u6e2c\u53ef\u4ee5\u7528\u7d71\u4e00\u6578\u64da\u96c6\u6bd4\u8f03\u3002<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">ConCor-1 \u55ba\u9577 caption \u6578\u64da\u96c6\u4e0a\u5c07 correspondence F1 \u63d0\u5347 48%\uff0c\u55ba\u96f6\u6a23\u672c LVIS\u3001\u5373\u7528\u5927\u985e\u5225\u6e05\u55ae\u7576\u6587\u5b57\u8f38\u5165\u5605\u5834\u666f\u4ea6\u63d0\u5347 29%\u3002Hugging Face \u4e0a\u5df2\u6709\u6a21\u578b\u6b0a\u91cd\u3001\u6578\u64da\u3001Space Demo\uff0c\u6e90\u4ee3\u78bc\u4ee5 Apache 2.0 \u767c\u4f48\u3002\u5c0d\u505a\u7d30\u7c92\u5ea6\u7406\u89e3\u3001\u81ea\u52d5\u5316\u6a19\u8a3b\uff0c\u6216\u8005\u60f3\u6574\u5408\u9577\u63cf\u8ff0\u5165\u8996\u89ba\u6d41\u7a0b\u5605\u5718\u968a\uff0c\u5462\u500b\u6846\u67b6\u503c\u5f97\u7559\u610f\uff1b\u60f3\u81ea\u884c\u91cd\u73fe\u8a13\u7df4\u5605\u4eba\u5c31\u8981\u518d\u7b49\u7b49\uff0c\u73fe\u968e\u6bb5\u4e3b\u8981\u63d0\u4f9b\u63a8\u8ad6\u7a0b\u5f0f\u78bc\u540c\u8a55\u6e2c\u5373\u5c07\u63a8\u51fa\u3002<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">\u91cd\u9ede\u6458\u8981<\/p>\n\n\n\n<ul class=\"wp-block-list\">\n<li>\u96d9\u5411\u6982\u5ff5\u5c0d\u61c9\uff1a\u6bcb\u9808\u9810\u5148\u6307\u660e\u6587\u5b57\uff0c\u76f4\u63a5\u8f38\u51fa\u6587\u5b57\u906e\u7f69\u3001\u5be6\u4f8b\u906e\u7f69\u540c\u5c0d\u61c9\u5206\u6578<\/li>\n<li>\u7d71\u4e00\u683c\u5f0f\uff1a\u5c07 phrase grounding\u3001referring expression\u3001open-vocabulary detection \u6536\u6210\u55ae\u4e00\u9810\u6e2c\u4efb\u52d9<\/li>\n<li>\u57fa\u65bc bridge tokens\uff1a\u55ba\u9810\u8a13\u7df4\u8996\u89ba\u8a9e\u8a00\u6a21\u578b\u4e0a\u52a0\u53ef\u5b78\u7fd2 token \u4ee3\u8868\u5019\u9078\u5c0d\u61c9<\/li>\n<li>\u986f\u8457\u63d0\u5347\uff1a\u9577 caption F1 \u63d0\u5347 48%\uff0c\u96f6\u6a23\u672c LVIS F1 \u63d0\u5347 29%<\/li>\n<li>\u958b\u6e90\u914d\u5957\uff1a\u6a21\u578b\u6b0a\u91cd\u3001\u6578\u64da\u3001Space Demo \u540c\u63a8\u8ad6\u7a0b\u5f0f\u78bc\u5df2\u65bc Hugging Face \u540c GitHub \u516c\u958b<\/li>\n<\/ul>\n\n\n\n<p class=\"wp-block-paragraph\">\u9069\u5408\u9700\u8981\u5f9e\u5716\u6587\u914d\u5c0d\u4e2d\u62bd\u53d6\u7d50\u69cb\u5316\u5c0d\u61c9\u3001\u505a\u81ea\u52d5\u5316\u6a19\u8a3b\u3001\u6216\u7814\u7a76\u8996\u89ba\u8a9e\u8a00\u5b9a\u4f4d\u6846\u67b6\u5605\u8b80\u8005\u3002ConCor-1 \u5c07\u6587\u5b57\u5206\u5272\u3001\u5f71\u50cf\u5206\u5272\u540c\u8de8\u6a21\u614b\u5c0d\u9f4a\u58d3\u6210\u55ae\u4e00\u4efb\u52d9\uff0c\u7279\u5225\u9069\u5408\u8655\u7406\u9577\u63cf\u8ff0\u6216\u5927\u985e\u5225\u6e05\u55ae\u7b49\u5834\u666f\u3002<\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><a href=\"https:\/\/uwgzq.github.io\/papers\/ConCor-1\/\" rel=\"noopener noreferrer\"><strong>\u9805\u76ee\u4e3b\u9801<\/strong><\/a> \u00b7 <a href=\"https:\/\/github.com\/RAIVNLab\/ConCor\" rel=\"noopener noreferrer\"><strong>GitHub<\/strong><\/a> \u00b7 <a href=\"https:\/\/huggingface.co\/UWGZQ\/ConCor-1\" rel=\"noopener noreferrer\"><strong>\u6a21\u578b<\/strong><\/a><\/p>\n","protected":false},"excerpt":{"rendered":"<p>ConCor-1 \u5c07\u8996\u89ba\u8a9e\u8a00\u5b9a\u4f4d\u53cd\u8f49\u6210\u96d9\u5411\u6982\u5ff5\u5c0d\u61c9\uff0c\u5514\u4f7f\u9810\u5148\u6307\u660e\u60f3\u6435\u4e5c\u5b57\u53e5\uff0c\u7540\u4e00\u5f35\u5716\u540c\u4e00\u6bb5\u6587\u5b57\u5c31\u6703\u81ea\u52d5\u5224\u5b9a\u908a\u5572\u6587\u5b57\u5c0d\u61c9\u908a\u500b\u7269\u4ef6\u3002<\/p>\n","protected":false},"author":8,"featured_media":10940,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"ai_generated_summary":"","footnotes":""},"categories":[133,119,76,149,188,199],"tags":[],"class_list":["post-10941","post","type-post","status-publish","format-standard","hentry","category-133","category-119","category-76","category-149","category-meta","category-dataset-"],"_links":{"self":[{"href":"https:\/\/infernews.com\/blog\/wp-json\/wp\/v2\/posts\/10941","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/infernews.com\/blog\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/infernews.com\/blog\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/infernews.com\/blog\/wp-json\/wp\/v2\/users\/8"}],"replies":[{"embeddable":true,"href":"https:\/\/infernews.com\/blog\/wp-json\/wp\/v2\/comments?post=10941"}],"version-history":[{"count":1,"href":"https:\/\/infernews.com\/blog\/wp-json\/wp\/v2\/posts\/10941\/revisions"}],"predecessor-version":[{"id":10943,"href":"https:\/\/infernews.com\/blog\/wp-json\/wp\/v2\/posts\/10941\/revisions\/10943"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/infernews.com\/blog\/wp-json\/wp\/v2\/media\/10940"}],"wp:attachment":[{"href":"https:\/\/infernews.com\/blog\/wp-json\/wp\/v2\/media?parent=10941"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/infernews.com\/blog\/wp-json\/wp\/v2\/categories?post=10941"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/infernews.com\/blog\/wp-json\/wp\/v2\/tags?post=10941"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}