{"id":8032,"date":"2026-05-07T00:33:37","date_gmt":"2026-05-06T16:33:37","guid":{"rendered":"https:\/\/infernews.com\/blog\/?p=8032"},"modified":"2026-05-07T00:34:35","modified_gmt":"2026-05-06T16:34:35","slug":"x2sam","status":"publish","type":"post","link":"https:\/\/infernews.com\/blog\/x2sam\/","title":{"rendered":"X2SAM\u628a\u5f71\u50cf\u8207\u5f71\u7247\u5206\u5272\u6574\u5408\u6210\u55ae\u4e00\u6a21\u578b"},"content":{"rendered":"<figure class=\"wp-block-image size-large\"><img decoding=\"async\" src=\"https:\/\/infernews.com\/blog\/wp-content\/uploads\/2026\/05\/framework-050a6973ff99.jpg\" alt=\"HuggingFace\" \/><\/figure>\n<p><a href=\"https:\/\/github.com\/wanghao9610\/X2SAM\">X2SAM<\/a> \u662f\u4e00\u500b\u7d71\u4e00\u5f0f\u5206\u5272\u591a\u6a21\u614b\u5927\u578b\u8a9e\u8a00\u6a21\u578b\uff0c\u76ee\u6a19\u662f\u628a\u5f71\u50cf\u4e2d\u7684\u300c\u4efb\u610f\u5206\u5272\u300d\u80fd\u529b\u5ef6\u4f38\u5230\u5f71\u7247\u3002\u5b83\u7d50\u5408 LLM\u3001Vision Encoder\u3001Mask Encoder\u3001Mask Decoder \u8207 Mask Memory\uff0c\u8b93\u6a21\u578b\u4e0d\u53ea\u7406\u89e3\u756b\u9762\u5167\u5bb9\uff0c\u9084\u80fd\u4f9d\u7167\u5c0d\u8a71\u6307\u4ee4\u6216\u8996\u89ba\u63d0\u793a\u7522\u751f\u50cf\u7d20\u7d1a\u906e\u7f69\u3002<\/p>\n<p>\u5be6\u969b\u4f7f\u7528\u4e0a\uff0c<a href=\"https:\/\/github.com\/wanghao9610\/X2SAM\">X2SAM<\/a> \u53ef\u540c\u6642\u63a5\u53d7<strong>\u5c0d\u8a71\u5f0f\u6587\u5b57\u6307\u4ee4<\/strong>\u8207<strong>\u8996\u89ba\u63d0\u793a<\/strong>\uff0c\u9069\u5408\u9700\u8981\u6307\u5b9a\u76ee\u6a19\u3001\u8ffd\u8e64\u7269\u4ef6\u6216\u4e92\u52d5\u4fee\u6b63\u7d50\u679c\u7684\u60c5\u5883\u3002\u5b98\u65b9\u63cf\u8ff0\u6307\u51fa\uff0c\u5b83\u652f\u63f4 generic\u3001open-vocabulary\u3001referring\u3001reasoning\u3001grounded conversation generation\u3001interactive \u8207 visual grounded segmentation\uff0c\u4ee3\u8868\u4f7f\u7528\u8005\u53ef\u7528\u8f03\u81ea\u7136\u7684\u65b9\u5f0f\u63d0\u51fa\u5206\u5272\u9700\u6c42\uff0c\u800c\u4e0d\u5fc5\u4fb7\u9650\u65bc\u55ae\u4e00\u8f38\u5165\u5f62\u5f0f\u3002<\/p>\n<p>\u9019\u500b\u5c08\u6848\u7684\u4e3b\u8981\u5275\u65b0\uff0c\u5728\u65bc\u7528\u55ae\u4e00\u4ecb\u9762\u6574\u5408\u5f71\u50cf\u8207\u5f71\u7247\u5206\u5272\uff0c\u4e26\u4ee5 <strong>Mask Memory<\/strong> \u5132\u5b58\u53d7\u5f15\u5c0e\u7684\u8996\u89ba\u7279\u5fb5\uff0c\u6539\u5584\u5f71\u7247\u4e2d\u8de8\u6642\u9593\u7684\u906e\u7f69\u4e00\u81f4\u6027\u3002\u6b64\u5916\uff0c\u4f5c\u8005\u4e5f\u63d0\u51fa <strong>V-VGD\uff08Video Visual Grounded\uff09<\/strong> \u5206\u5272\u57fa\u6e96\uff0c\u7528\u4f86\u8a55\u4f30\u6a21\u578b\u662f\u5426\u80fd\u6839\u64da\u4e92\u52d5\u5f0f\u8996\u89ba\u63d0\u793a\uff0c\u5728\u5f71\u7247\u4e2d\u5206\u5272\u4e26\u8ffd\u8e64\u7269\u4ef6\u3002<\/p>\n<ul>\n<li>\u7d71\u4e00\u652f\u63f4\u5f71\u50cf\u8207\u5f71\u7247\u5206\u5272\uff0c\u800c\u975e\u53ea\u5c08\u6ce8\u55ae\u4e00\u5a92\u9ad4<\/li>\n<li>\u540c\u6642\u652f\u63f4\u6587\u5b57\u6307\u4ee4\u8207\u8996\u89ba\u63d0\u793a\u8f38\u5165<\/li>\n<li>\u900f\u904e Mask Memory \u5f37\u5316\u5f71\u7247\u906e\u7f69\u7684\u6642\u5e8f\u4e00\u81f4\u6027<\/li>\n<li>\u63d0\u51fa V-VGD \u57fa\u6e96\u88dc\u8db3\u5f71\u7247\u8996\u89ba\u5b9a\u4f4d\u5206\u5272\u8a55\u4f30<\/li>\n<li>\u63a1\u7528\u7570\u8cea\u5f71\u50cf\u8207\u5f71\u7247\u8cc7\u6599\u7684\u806f\u5408\u8a13\u7df4\u7b56\u7565<\/li>\n<\/ul>\n<p>\u5f9e\u5de5\u4f5c\u61c9\u7528\u4f86\u770b\uff0c\u9019\u985e\u7cfb\u7d71\u53ef\u671b\u53d7\u60e0\u65bc\u5f71\u7247\u5167\u5bb9\u7406\u89e3\u3001\u4e92\u52d5\u5f0f\u6a19\u8a3b\u3001\u667a\u6167\u526a\u8f2f\u3001\u8996\u89ba\u52a9\u7406\u8207\u591a\u6a21\u614b\u4eba\u6a5f\u4e92\u52d5\u7b49\u4efb\u52d9\u3002\u6027\u80fd\u65b9\u9762\uff0c\u539f\u6587\u8868\u793a X2SAM \u5728\u5f71\u7247\u5206\u5272\u4e0a\u9054\u5230\u5f37\u52c1\u8868\u73fe\uff0c\u5c0d\u5f71\u50cf\u5206\u5272\u57fa\u6e96\u4ecd\u5177\u7af6\u722d\u529b\uff0c\u4e26\u4fdd\u7559\u4e00\u822c\u5f71\u50cf\u8207\u5f71\u7247\u804a\u5929\u80fd\u529b\uff1b\u4e0d\u904e\u6b64\u9801\u9762\u672a\u5b8c\u6574\u5217\u51fa\u5177\u9ad4\u6578\u503c\uff0c\u56e0\u6b64\u89e3\u8b80\u4e0a\u4ecd\u61c9\u4ee5\u8ad6\u6587\u8207\u5be6\u9a57\u8868\u683c\u70ba\u6e96\u3002<\/p>\n<p>\u6a21\u578b\u5217\u8868\uff1aLLM\u3001SAM \u7cfb\u5217<\/p>\n","protected":false},"excerpt":{"rendered":"<p>X2SAM\u5c07\u5f71\u50cf\u8207\u5f71\u7247\u5206\u5272\u6574\u5408\u5230\u540c\u4e00\u5957\u591a\u6a21\u614b\u67b6\u69cb\u4e2d\uff0c\u652f\u63f4\u6587\u5b57\u6307\u4ee4\u8207\u8996\u89ba\u63d0\u793a\u3002\u5b83\u4e5f\u5f37\u8abf\u5f71\u7247\u906e\u7f69\u7684\u4e00\u81f4\u6027\u8207\u8de8\u4efb\u52d9\u901a\u7528\u6027\u3002<\/p>\n","protected":false},"author":8,"featured_media":8031,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"googlesitekit_rrm_CAowvqSiDA:productID":"","footnotes":""},"categories":[133,157,120],"tags":[],"class_list":["post-8032","post","type-post","status-publish","format-standard","hentry","category-133","category-157","category-120"],"_links":{"self":[{"href":"https:\/\/infernews.com\/blog\/wp-json\/wp\/v2\/posts\/8032","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=8032"}],"version-history":[{"count":2,"href":"https:\/\/infernews.com\/blog\/wp-json\/wp\/v2\/posts\/8032\/revisions"}],"predecessor-version":[{"id":8035,"href":"https:\/\/infernews.com\/blog\/wp-json\/wp\/v2\/posts\/8032\/revisions\/8035"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/infernews.com\/blog\/wp-json\/wp\/v2\/media\/8031"}],"wp:attachment":[{"href":"https:\/\/infernews.com\/blog\/wp-json\/wp\/v2\/media?parent=8032"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/infernews.com\/blog\/wp-json\/wp\/v2\/categories?post=8032"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/infernews.com\/blog\/wp-json\/wp\/v2\/tags?post=8032"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}