{"id":12171,"date":"2023-01-30T20:25:37","date_gmt":"2023-01-30T12:25:37","guid":{"rendered":"http:\/\/139.9.1.231\/?p=12171"},"modified":"2023-01-31T09:57:18","modified_gmt":"2023-01-31T01:57:18","slug":"monocular-depth-estimation","status":"publish","type":"post","link":"http:\/\/139.9.1.231\/index.php\/2023\/01\/30\/monocular-depth-estimation\/","title":{"rendered":"\u57fa\u4e8e\u6df1\u5ea6\u5b66\u4e60\u7684\u5355\u76ee\u6df1\u5ea6\u4f30\u8ba1\u7efc\u8ff0"},"content":{"rendered":"\n<p class=\"has-text-align-center has-light-pink-background-color has-background\"><strong><a href=\"https:\/\/paperswithcode.com\/task\/monocular-depth-estimation\" target=\"_blank\" rel=\"noreferrer noopener\">Monocular Depth Estimation<\/a><\/strong><\/p>\n\n\n\n<p><strong>Monocular Depth Estimation<\/strong> is the task of estimating the depth value (distance relative to the camera) of each pixel given a single (monocular) RGB image. This challenging task is a key prerequisite for determining scene understanding for applications such as 3D scene reconstruction, autonomous driving, and AR. State-of-the-art methods usually fall into one of two categories: designing a complex network that is powerful enough to directly regress the depth map, or splitting the input into bins or windows to reduce computational complexity. The most popular benchmarks are the KITTI and NYUv2 datasets. Models are typically evaluated using RMSE or absolute relative error. <strong><em>\u8fd9\u9879\u5177\u6709\u6311\u6218\u6027\u7684\u4efb\u52a1\u662f\u786e\u5b9a 3D \u573a\u666f\u91cd\u5efa\u3001\u81ea\u52a8\u9a7e\u9a76\u548c AR \u7b49\u5e94\u7528\u573a\u666f\u7406\u89e3\u7684\u5173\u952e\u5148\u51b3\u6761\u4ef6\u3002<\/em><\/strong><\/p>\n\n\n\n<h3><strong>\u4efb\u52a1\u4ecb\u7ecd<\/strong><\/h3>\n\n\n\n<p>\u6df1\u5ea6\u4f30\u8ba1\u662f\u8ba1\u7b97\u673a\u89c6\u89c9\u9886\u57df\u7684\u4e00\u4e2a\u57fa\u7840\u6027\u95ee\u9898\uff0c\u5176\u53ef\u4ee5\u5e94\u7528\u5728\u673a\u5668\u4eba\u5bfc\u822a\u3001\u589e\u5f3a\u73b0\u5b9e\u3001\u4e09\u7ef4\u91cd\u5efa\u3001\u81ea\u52a8\u9a7e\u9a76\u7b49\u9886\u57df\u3002\u800c\u76ee\u524d\u5927\u90e8\u5206\u6df1\u5ea6\u4f30\u8ba1\u90fd\u662f\u57fa\u4e8e\u4e8c\u7ef4RGB\u56fe\u50cf\u5230RBG-D\u56fe\u50cf\u7684\u8f6c\u5316\u4f30\u8ba1\uff0c\u4e3b\u8981\u5305\u62ec\u4ece\u56fe\u50cf\u660e\u6697\u3001\u4e0d\u540c\u89c6\u89d2\u3001\u5149\u5ea6\u3001\u7eb9\u7406\u4fe1\u606f\u7b49\u83b7\u53d6\u573a\u666f\u6df1\u5ea6\u5f62\u72b6\u7684Shape from X\u65b9\u6cd5\uff0c\u8fd8\u6709\u7ed3\u5408SFM(Structure from motion)\u548cSLAM(Simultaneous Localization And Mapping)\u7b49\u65b9\u5f0f\u9884\u6d4b\u76f8\u673a\u4f4d\u59ff\u7684\u7b97\u6cd5\u3002\u5176\u4e2d\u867d\u7136\u6709\u5f88\u591a\u8bbe\u5907\u53ef\u4ee5\u76f4\u63a5\u83b7\u53d6\u6df1\u5ea6\uff0c\u4f46\u662f\u8bbe\u5907\u9020\u4ef7\u6602\u8d35\u3002\u4e5f\u53ef\u4ee5\u5229\u7528\u53cc\u76ee\u8fdb\u884c\u6df1\u5ea6\u4f30\u8ba1\uff0c\u4f46\u662f\u7531\u4e8e\u53cc\u76ee\u56fe\u50cf\u9700\u8981\u5229\u7528\u7acb\u4f53\u5339\u914d\u8fdb\u884c\u50cf\u7d20\u70b9\u5bf9\u5e94\u548c\u89c6\u5dee\u8ba1\u7b97\uff0c\u6240\u4ee5\u8ba1\u7b97\u590d\u6742\u5ea6\u4e5f\u8f83\u9ad8\uff0c\u5c24\u5176\u662f\u5bf9\u4e8e\u4f4e\u7eb9\u7406\u573a\u666f\u7684\u5339\u914d\u6548\u679c\u4e0d\u597d\u3002\u800c\u5355\u76ee\u6df1\u5ea6\u4f30\u8ba1\u5219\u76f8\u5bf9\u6210\u672c\u66f4\u4f4e\uff0c\u66f4\u5bb9\u6613\u666e\u53ca\u3002<\/p>\n\n\n\n<p>\u90a3\u4e48\u5bf9\u4e8e\u5355\u76ee\u6df1\u5ea6\u4f30\u8ba1\uff0c\u987e\u540d\u601d\u4e49\uff0c\u5c31\u662f\u5229\u7528\u4e00\u5f20\u6216\u8005\u552f\u4e00\u89c6\u89d2\u4e0b\u7684RGB\u56fe\u50cf\uff0c\u4f30\u8ba1\u56fe\u50cf\u4e2d\u6bcf\u4e2a\u50cf\u7d20\u76f8\u5bf9\u62cd\u6444\u6e90\u7684\u8ddd\u79bb\u3002\u5bf9\u4e8e\u4eba\u773c\u6765\u8bf4\uff0c\u7531\u4e8e\u5b58\u5728\u5927\u91cf\u7684\u5148\u9a8c\u77e5\u8bc6\uff0c\u6240\u4ee5\u53ef\u4ee5\u4ece\u4e00\u53ea\u773c\u775b\u6240\u83b7\u53d6\u7684\u56fe\u50cf\u4fe1\u606f\u4e2d\u63d0\u53d6\u51fa\u5927\u91cf\u6df1\u5ea6\u4fe1\u606f\u3002\u90a3\u4e48\u5355\u76ee\u6df1\u5ea6\u4f30\u8ba1\u4e0d\u4ec5\u9700\u8981\u4ece\u4e8c\u7ef4\u56fe\u50cf\u4e2d\u5b66\u4f1a\u5ba2\u89c2\u7684\u6df1\u5ea6\u4fe1\u606f\uff0c\u800c\u4e14\u9700\u8981\u63d0\u53d6\u4e00\u4e9b\u7ecf\u9a8c\u4fe1\u606f\uff0c\u540e\u8005\u5219\u5bf9\u4e8e\u6570\u636e\u96c6\u4e2d\u76f8\u673a\u548c\u573a\u666f\u4f1a\u6bd4\u8f83\u654f\u611f\u3002<\/p>\n\n\n\n<p>\u901a\u8fc7\u9605\u8bfb\u6587\u732e\uff0c\u53ef\u4ee5\u5c06\u57fa\u4e8e\u6df1\u5ea6\u5b66\u4e60\u7684\u5355\u76ee\u6df1\u5ea6\u4f30\u8ba1\u7b97\u6cd5\u5927\u81f4\u5206\u4e3a\u4ee5\u4e0b\u51e0\u7c7b\uff1a<\/p>\n\n\n\n<ul><li><strong>\u76d1\u7763\u7b97\u6cd5<\/strong><\/li><\/ul>\n\n\n\n<p>\u987e\u540d\u601d\u4e49\uff0c\u76f4\u63a5\u4ee52\u7ef4\u56fe\u50cf\u4f5c\u4e3a\u8f93\u5165\uff0c\u4ee5\u6df1\u5ea6\u56fe\u4e3a\u8f93\u51fa\u8fdb\u884c\u8bad\u7ec3\uff1a\uff1a\u76d1\u7763\u65b9\u6cd5\u7684\u76d1\u7763\u4fe1\u53f7\u57fa\u4e8e\u6df1\u5ea6\u56fe\u7684\u5730\u9762\u771f\u503c\uff0c\u56e0\u6b64\u5355\u76ee\u6df1\u5ea6\u4f30\u8ba1\u53ef\u4ee5\u770b\u4f5c\u662f\u4e00\u4e2a\u56de\u5f52\u95ee\u9898\u3002\u4ece\u5355\u4e2a\u6df1\u5ea6\u56fe\u50cf\u8bbe\u8ba1\u795e\u7ecf\u7f51\u7edc\u6765\u9884\u6d4b\u6df1\u5ea6\u3002\u5229\u7528\u9884\u6d4b\u6df1\u5ea6\u56fe\u548c\u5b9e\u9645\u6df1\u5ea6\u56fe\u4e4b\u95f4\u7684\u5dee\u5f02\u6765\u76d1\u7763\u7f51\u7edc\u7684\u8bad\u7ec3 L2\u635f\u5931<\/p>\n\n\n\n<div class=\"wp-block-image\"><figure class=\"aligncenter\"><img src=\"https:\/\/pic4.zhimg.com\/v2-20187e2105287415219df37e8f6fea3b_r.jpg\" alt=\"\"\/><\/figure><\/div>\n\n\n\n<div class=\"wp-block-image\"><figure class=\"aligncenter\"><img src=\"https:\/\/pic2.zhimg.com\/v2-393111af6b376f173836c7cf594b3aad_r.jpg\" alt=\"\"\/><\/figure><\/div>\n\n\n\n<p>\u4e0a\u9762\u7ed9\u7684\u4f8b\u5b50\u662fKITTI\u6570\u636e\u96c6\u4e2d\u7684\u4e00\u7ec4\u4f8b\u5b50\uff0c\u4e0d\u8fc7\u6df1\u5ea6\u56fe\u53ef\u80fd\u770b\u7684\u4e0d\u662f\u5f88\u660e\u663e\uff0c\u6211\u91cd\u65b0\u5c06\u6df1\u5ea6\u56fe\u6d82\u8272\u4e4b\u540e\uff1a<\/p>\n\n\n\n<div class=\"wp-block-image\"><figure class=\"aligncenter\"><img src=\"https:\/\/pic2.zhimg.com\/v2-01bc9fff2f8e84950ceee5df7c4ee7c9_r.jpg\" alt=\"\"\/><\/figure><\/div>\n\n\n\n<p>       \u6df1\u5ea6\u7f51\u7edc\u901a\u8fc7\u8fd1\u4f3c\u771f\u503c\u7684\u65b9\u6cd5\u6765\u5b66\u4e60\u573a\u666f\u7684\u6df1\u5ea6\u3002\u57fa\u4e8e\u4e0d\u540c\u7ed3\u6784\u548c\u635f\u5931\u51fd\u6570\u7684\u65b9\u6cd5\uff1a\u636e\u6211\u4eec\u6240\u77e5\uff0cEigen\u7b49\u4eba\u9996\u5148\u7528CNNs\u89e3\u51b3\u5355\u76ee\u6df1\u5ea6\u4f30\u8ba1\u95ee\u9898\u3002\u8be5\u4f53\u7cfb\u7ed3\u6784\u7531\u4e24\u4e2a\u7ec4\u6210\u90e8\u5206\u7ec4\u6210\uff08\u5168\u5c40\u7c97\u5c3a\u5ea6\u7f51\u7edc\u548c\u5c40\u90e8\u7cbe\u7ec6\u5c3a\u5ea6\u7f51\u7edc\uff09\uff0c\u5728\u6587\u732e\u4e2d\u7528\u4e8e\u4ece\u5355\u4e2a\u56fe\u50cf\u8fdb\u884c\u7aef\u5230\u7aef\u7684\u6df1\u5ea6\u56fe\u9884\u6d4b\u3002<\/p>\n\n\n\n<p>       \u57fa\u4e8e\u6761\u4ef6\u968f\u673a\u573a\u7684\u65b9\u6cd5\uff1aLi\u7b49\u4eba\u63d0\u51fa\u4e86\u4e00\u79cd\u57fa\u4e8e\u591a\u5c42\u7684\u6761\u4ef6\u968f\u673a\u573a\uff08CRFs\uff09\u7684\u7ec6\u5316\u65b9\u6cd5\uff0c\u8be5\u65b9\u6cd5\u4e5f\u88ab\u5e7f\u6cdb\u5e94\u7528\u4e8e\u8bed\u4e49\u5206\u5272\u3002\u5728\u6df1\u5ea6\u7684\u4f30\u8ba1\u4e2d\uff0c\u8003\u8651\u5230\u6df1\u5ea6\u7684\u8fde\u7eed\u7279\u5f81\uff0c\u53ef\u4ee5\u5e7f\u6cdb\u5730\u4f7f\u7528CRF\u7684\u6df1\u5ea6\u4fe1\u606f\uff0c\u56e0\u6b64\u53ef\u4ee5\u5e7f\u6cdb\u5730\u5e94\u7528\u4e8e\u6df1\u5ea6\u7684\u4f30\u8ba1\u4e2d\u3002<\/p>\n\n\n\n<p>\u57fa\u4e8e\u5bf9\u6297\u6027\u5b66\u4e60\u7684\u65b9\u6cd5\uff1a\u7531\u4e8e\u63d0\u51fa\u7684\u5bf9\u6297\u6027\u5b66\u4e60\u5728\u6570\u636e\u751f\u6210\u65b9\u9762\u7684\u7a81\u51fa\u8868\u73b0\uff0c\u8fd1\u5e74\u6765\u6210\u4e3a\u4e00\u4e2a\u7814\u7a76\u70ed\u70b9\u3002\u5404\u79cd\u7b97\u6cd5\u3001\u7406\u8bba\u548c\u5e94\u7528\u5df2\u5f97\u5230\u5e7f\u6cdb\u53d1\u5c55\u3002\u5bf9\u6297\u5f0f\u5b66\u4e60\u6df1\u5ea6\u4f30\u8ba1\u7684\u6846\u67b6\u5982\u56fe\u6240\u793a\u3002<\/p>\n\n\n\n<figure class=\"wp-block-image\"><img src=\"https:\/\/minio.cvmart.net\/cvmart-community\/images\/202010\/10\/73\/5tbYHDOi4P.png?imageView2\/2\/w\/1240\/h\/0\" alt=\"file\"\/><\/figure>\n\n\n\n<ul><li><strong>\u65e0\u76d1\u7763\u7b97\u6cd5<\/strong><\/li><\/ul>\n\n\n\n<p><strong><em>\u9996\u5148\uff0c\u6240\u8c13\u7684\u201c\u65e0\u76d1\u7763\u201d\u867d\u7136\u4e0d\u9700\u8981\u8f93\u5165\u771f\u5b9e\u6df1\u5ea6\u4fe1\u606f\uff0c\u4f46\u9700\u8981\u8f93\u5165\u53cc\u76ee\u6444\u50cf\u5934\u83b7\u53d6\u5230\u7684\u540c\u4e00\u65f6\u523b\u4e0d\u540c\u89d2\u5ea6\u7684\u56fe\u50cf\u6216\u8005\u524d\u540e\u5e27\u56fe\u50cf\uff0c\u53ea\u662f\u8fd9\u6837\u5c31\u53eb\u505a\u65e0\u76d1\u7763\u5728\u6211\u770b\u6765\u7565\u663e\u7275\u5f3a\u3002<\/em><\/strong><\/p>\n\n\n\n<p>\u6709\u76d1\u7763\u5b66\u4e60\u65b9\u6cd5\u8981\u6c42\u6bcf\u5e45RGB\u56fe\u50cf\u90fd\u6709\u5176\u5bf9\u5e94\u7684\u6df1\u5ea6\u6807\u7b7e\uff0c\u800c\u6df1\u5ea6\u6807\u7b7e\u91c7\u96c6\u901a\u5e38\u9700\u8981\u6df1\u5ea6\u76f8\u673a\u6216\u6fc0\u5149\u96f7\u8fbe\uff0c\u524d\u8005\u8303\u56f4\u53d7\u9650\u540e\u8005\u6210\u672c\u6602\u8d35\u3002\u518d\u8005\uff0c\u91c7\u96c6\u7684\u539f\u59cb\u6df1\u5ea6\u6807\u7b7e\u901a\u5e38\u662f\u4e00\u4e9b\u7a00\u758f\u7684\u70b9\uff0c\u4e0d\u80fd\u4e0e\u539f\u56fe\u5f88\u597d\u7684\u5339\u914d\u3002\u56e0\u6b64\u4e0d\u7528\u6df1\u5ea6\u6807\u7b7e\u7684\u65e0\u76d1\u7763\u4f30\u8ba1\u65b9\u6cd5\u662f\u8fd1\u5e74\u7684\u7814\u7a76\u8d8b\u52bf\uff0c<strong>\u5176\u57fa\u672c\u601d\u8def\u662f\u5229\u7528\u5de6\u53f3\u89c6\u56fe\uff0c\u7ed3\u5408\u5bf9\u6781\u51e0\u4f55\u4e0e\u81ea\u52a8\u7f16\u7801\u673a\u7684\u601d\u60f3\u6c42\u89e3\u6df1\u5ea6\u3002&nbsp;<\/strong><\/p>\n\n\n\n<p>\u7531\u4e8e\u6df1\u5ea6\u6570\u636e\u7684\u83b7\u53d6\u96be\u5ea6\u8f83\u9ad8\uff0c\u6240\u4ee5\u76ee\u524d\u6709\u5927\u91cf\u7b97\u6cd5\u90fd\u662f\u57fa\u4e8e\u65e0\u76d1\u7763\u6a21\u578b\u7684\u3002\u5373\u4ec5\u4ec5\u4f7f\u7528\u4e24\u4e2a\u6444\u50cf\u673a\u91c7\u96c6\u7684\u53cc\u76ee\u56fe\u50cf\u6570\u636e\u8fdb\u884c\u8054\u5408\u8bad\u7ec3\u3002<strong>\u5176\u4e2d\u53cc\u76ee\u6570\u636e\u53ef\u5f7c\u6b64\u9884\u6d4b\u5bf9\u65b9\uff0c\u4ece\u800c\u83b7\u5f97\u76f8\u5e94\u7684\u89c6\u5dee\u6570\u636e\uff0c\u518d\u6839\u636e\u89c6\u5dee\u4e0e\u6df1\u5ea6\u7684\u5173\u7cfb\u8fdb\u884c\u6f14\u5316\u3002\u4ea6\u6216\u662f\u5c06\u53cc\u76ee\u56fe\u50cf\u4e2d\u5404\u4e2a\u50cf\u7d20\u70b9\u7684\u5bf9\u5e94\u95ee\u9898\u770b\u4f5c\u662f\u7acb\u4f53\u5339\u914d\u95ee\u9898\u8fdb\u884c\u8bad\u7ec3\u3002<\/strong>\u5de6\u89c6\u56fe-\u53f3\u89c6\u56fe\u793a\u4f8b\uff1a<\/p>\n\n\n\n<div class=\"wp-block-image\"><figure class=\"aligncenter\"><img src=\"https:\/\/pic4.zhimg.com\/v2-492dd0dc14c667d06ff41ed36c5ee5af_r.jpg\" alt=\"\"\/><\/figure><\/div>\n\n\n\n<div class=\"wp-block-image\"><figure class=\"aligncenter\"><img src=\"https:\/\/pic2.zhimg.com\/v2-c571282f8740f9013fc8dc2b2857c709_r.jpg\" alt=\"\"\/><\/figure><\/div>\n\n\n\n<p>\u89c6\u5dee\uff0c\u4ee5\u6211\u4eec\u4eba\u773c\u4e3a\u4f8b\uff0c\u4e24\u53ea\u773c\u775b\u770b\u5230\u7684\u56fe\u50cf\u5206\u522b\u4f4d\u4e8e\u4e0d\u540c\u7684\u5750\u6807\u7cfb\u3002\u5c06\u624b\u6307\u4ece\u8f83\u8fdc\u5730\u65b9\u6162\u6162\u79fb\u52a8\u5230\u773c\u524d\uff0c\u4f1a\u53d1\u73b0\uff0c\u624b\u6307\u5728\u5de6\u773c\u7684\u5750\u6807\u7cfb\u4e2d\u8d8a\u6765\u8d8a\u9760\u53f3\uff0c\u800c\u5728\u53f3\u773c\u5750\u6807\u7cfb\u4e2d\u8d8a\u6765\u8d8a\u9760\u5de6\uff0c\u8fd9\u79cd\u5dee\u5f02\u6027\u5c31\u662f\u89c6\u5dee\u3002\u4e0e\u6b64\u540c\u65f6\uff0c\u53ef\u4ee5\u8bf4\u660e\uff0c\u89c6\u5dee\u4e0e\u6df1\u5ea6\u6210\u53cd\u6bd4\u3002\u9664\u6b64\u4e4b\u5916\uff0c\u7531\u4e8e\u6444\u50cf\u673a\u53c2\u6570\u4e5f\u6bd4\u8f83\u5bb9\u6613\u83b7\u53d6\uff0c\u6240\u4ee5\u4e5f\u53ef\u4ee5\u4ee5\u76f8\u673a\u4f4d\u59ff\u4f5c\u4e3a\u6807\u7b7e\u8fdb\u884c\u8bad\u7ec3\u3002<\/p>\n\n\n\n<div class=\"wp-block-image\"><figure class=\"aligncenter\"><img src=\"https:\/\/pic4.zhimg.com\/80\/v2-7efbfc80cdb24ebc0428cd11a903036b_1440w.webp\" alt=\"\"\/><\/figure><\/div>\n\n\n\n<p>\u540c\u65f6\u540c\u4e00\u6c34\u5e73\u7ebf\u4e0a\u7684\u4e24\u4e2a\u7167\u76f8\u673a\u62cd\u6444\u5230\u7684\u7167\u7247\u662f\u670d\u4ece\u4ee5\u4e0b\u7269\u7406\u89c4\u5f8b\u7684\uff1a<\/p>\n\n\n\n<figure class=\"wp-block-image\"><img src=\"https:\/\/pic3.zhimg.com\/v2-8160f8817f689bcfcd0b287b05b72c2a_r.jpg\" alt=\"\"\/><\/figure>\n\n\n\n<p>\u5728\u56fe\u4e2d\uff0c&nbsp;Z&nbsp;\u4e3a\u573a\u666f\u6240\u8ddd\u79bb\u6211\u4eec\u7684\u6df1\u5ea6,&nbsp;X\u4e3a\u4e09\u7ef4\u573a\u666f\u6620\u5c04\u5230\u7684\u4e8c\u7ef4\u56fe\u50cf\u5e73\u9762\uff0c\u4e5f\u5c31\u662f\u6700\u7ec8\u6211\u4eec\u5f97\u5230\u7684\u4e8c\u7ef4\u56fe\u50cf\u6240\u5728\u7684\u5e73\u9762\u3002&nbsp;f\u4e3a\u76f8\u673a\u7684\u7126\u8ddd\u3002&nbsp;b\u4e3a\u4e24\u4e2a\u76f8\u673a\u4e4b\u95f4\u7684\u8ddd\u79bb\uff0cXl\u548c&nbsp;Xr&nbsp;\u5206\u522b\u4e3a\u76f8\u540c\u7269\u4f53\u5728\u5de6\u53f3\u4e24\u4e2a\u4e0d\u540c\u76f8\u673a\u4e2d\u6210\u50cf\u7684\u5750\u6807\u3002\u6839\u636e\u4ee5\u4e0a\u4fe1\u606f\uff0c\u548c\u7b80\u5355\u7684\u4e09\u89d2\u5f62\u76f8\u4f3c\u89c4\u5f8b\u6211\u4eec\u53ef\u4ee5\u5f97\u5230\uff1a<\/p>\n\n\n\n<figure class=\"wp-block-image size-large is-resized\"><img loading=\"lazy\" src=\"http:\/\/139.9.1.231\/wp-content\/uploads\/2023\/01\/image-83-1024x429.png\" alt=\"\" class=\"wp-image-12196\" width=\"690\" height=\"289\" srcset=\"http:\/\/139.9.1.231\/wp-content\/uploads\/2023\/01\/image-83-1024x429.png 1024w, http:\/\/139.9.1.231\/wp-content\/uploads\/2023\/01\/image-83-300x126.png 300w, http:\/\/139.9.1.231\/wp-content\/uploads\/2023\/01\/image-83-768x322.png 768w, http:\/\/139.9.1.231\/wp-content\/uploads\/2023\/01\/image-83.png 1110w\" sizes=\"(max-width: 690px) 100vw, 690px\" \/><\/figure>\n\n\n\n<figure class=\"wp-block-image size-large\"><img loading=\"lazy\" width=\"1024\" height=\"308\" src=\"http:\/\/139.9.1.231\/wp-content\/uploads\/2023\/01\/image-84-1024x308.png\" alt=\"\" class=\"wp-image-12197\" srcset=\"http:\/\/139.9.1.231\/wp-content\/uploads\/2023\/01\/image-84-1024x308.png 1024w, http:\/\/139.9.1.231\/wp-content\/uploads\/2023\/01\/image-84-300x90.png 300w, http:\/\/139.9.1.231\/wp-content\/uploads\/2023\/01\/image-84-768x231.png 768w, http:\/\/139.9.1.231\/wp-content\/uploads\/2023\/01\/image-84.png 1125w\" sizes=\"(max-width: 1024px) 100vw, 1024px\" \/><\/figure>\n\n\n\n<p>\u8fd9\u79cd\u601d\u8def\u6700\u5148\u5e94\u7528\u4e8e\u4f7f\u7528\u5355\u5f20\u56fe\u7247\u751f\u6210\u65b0\u89c6\u89d2\u95ee\u9898\uff1a<a rel=\"noreferrer noopener\" href=\"https:\/\/arxiv.org\/abs\/1506.06825\" target=\"_blank\">DeepStereo<\/a>&nbsp;\u548c&nbsp;<a rel=\"noreferrer noopener\" href=\"https:\/\/arxiv.org\/abs\/1604.03650\" target=\"_blank\">Deep3d<\/a>\u4e4b\u4e2d, \u5728\u4f20\u7edf\u7684\u89c6\u89d2\u751f\u6210\u95ee\u9898\u4e4b\u4e2d\uff0c\u9996\u5148\u4f1a\u5229\u7528\u4e24\u5f20\u56fe\uff08\u6216\u591a\u5f20\uff09\u6c42\u53d6\u56fe\u7247\u4e4b\u95f4\u7684\u89c6\u5deed\uff0c\u5176\u6b21\u901a\u8fc7\u5f97\u5230\u7684\u89c6\u5dee\uff08\u76f8\u5f53\u4e8e\u4e09\u7ef4\u573a\u666f\uff09\u6765\u751f\u6210\u65b0\u89c6\u89d2\u3002<\/p>\n\n\n\n<p>\u57fa\u4e8e\u53ef\u89e3\u91ca\u6027\u63a9\u6a21\u7684\u65b9\u6cd5\uff1a\u57fa\u4e8e\u6295\u5f71\u51fd\u6570\u7684\u89c6\u56fe\u91cd\u5efa\u7b97\u6cd5\u4f9d\u8d56\u4e8e\u9759\u6001\u573a\u666f\u5047\u8bbe\uff0c\u5373\u52a8\u6001\u76ee\u6807\u5728\u76f8\u90bb\u5e27\u4e0a\u7684\u4f4d\u7f6e\u4e0d\u6ee1\u8db3\u6295\u5f71\u51fd\u6570\uff0c\u4ece\u800c\u5f71\u54cd\u6d4b\u5149\u5ea6\u8bef\u5dee\u548c\u8bad\u7ec3\u8fc7\u7a0b\u3002<\/p>\n\n\n\n<p>\u57fa\u4e8e\u4f20\u7edf\u89c6\u89c9\u91cc\u7a0b\u8ba1\u7684\u65b9\u6cd5\uff1a\u7528\u4f20\u7edf\u7684\u76f4\u63a5\u89c6\u89c9\u91cc\u7a0b\u8ba1\u56de\u5f52\u7684\u4f4d\u59ff\u6765\u8f85\u52a9\u6df1\u5ea6\u4f30\u8ba1\uff0c\u800c\u4e0d\u662f\u4f7f\u7528\u4f4d\u59ff\u7f51\u7edc\u4f30\u8ba1\u7684\u4f4d\u59ff\u3002\u76f4\u63a5\u89c6\u89c9\u91cc\u7a0b\u8ba1\u5229\u7528\u6df1\u5ea6\u7f51\u7edc\u751f\u6210\u7684\u6df1\u5ea6\u56fe\u548c\u4e00\u4e2a\u4e09\u5e27\u56fe\u50cf\uff0c\u901a\u8fc7\u6700\u5c0f\u5316\u5149\u5ea6\u8bef\u5dee\u6765\u4f30\u8ba1\u5e27\u95f4\u7684\u59ff\u6001\uff0c\u7136\u540e\u5c06\u8ba1\u7b97\u51fa\u7684\u59ff\u6001\u53d1\u9001\u56de\u8bad\u7ec3\u6846\u67b6\u3002\u56e0\u6b64\uff0c\u7531\u4e8e\u6df1\u5ea6\u7f51\u7edc\u7531\u66f4\u7cbe\u786e\u7684\u59ff\u6001\u6765\u76d1\u7763\uff0c\u56e0\u6b64\u6df1\u5ea6\u4f30\u8ba1\u7684\u7cbe\u5ea6\u663e\u7740\u63d0\u9ad8\u3002<\/p>\n\n\n\n<p>\u57fa\u4e8e\u591a\u4efb\u52a1\u6846\u67b6\u7684\u65b9\u6cd5\uff1a\u6700\u8fd1\u7684\u65b9\u6cd5\u5728\u57fa\u672c\u6846\u67b6\u4e2d\u5f15\u5165\u4e86\u989d\u5916\u7684\u591a\u4efb\u52a1\u7f51\u7edc\uff0c\u5982\u5149\u6d41\u3001\u7269\u4f53\u8fd0\u52a8\u548c\u76f8\u673a\u5185\u53c2\u77e9\u9635\uff0c\u4f5c\u4e3a\u4e00\u4e2a\u9644\u52a0\u7684\u8bad\u7ec3\u6846\u67b6\uff0c\u52a0\u5f3a\u4e86\u6574\u4e2a\u8bad\u7ec3\u4efb\u52a1\u4e4b\u95f4\u7684\u5173\u7cfb<\/p>\n\n\n\n<p>\u57fa\u4e8e\u5bf9\u6297\u5b66\u4e60\u7684\u65b9\u6cd5\uff1a\u5c06\u5bf9\u6297\u5b66\u4e60\u6846\u67b6\u5f15\u5165\u5230\u65e0\u76d1\u7763\u7684\u5355\u76ee\u6df1\u5ea6\u4f30\u8ba1\u4e2d\u3002\u7531\u4e8e\u5728\u65e0\u76d1\u7763\u8bad\u7ec3\u4e2d\u6ca1\u6709\u771f\u6b63\u7684\u6df1\u5ea6\u56fe\u3002\u56e0\u6b64\uff0c\u5c06\u89c6\u56fe\u91cd\u5efa\u7b97\u6cd5\u5408\u6210\u7684\u56fe\u50cf\u548c\u771f\u5b9e\u56fe\u50cf\u4f5c\u4e3a\u9274\u522b\u5668\u7684\u8f93\u5165\uff0c\u800c\u4e0d\u662f\u4f7f\u7528\u9274\u522b\u5668\u6765\u533a\u5206\u771f\u5b9e\u6df1\u5ea6\u56fe\u548c\u9884\u6d4b\u6df1\u5ea6\u56fe\u3002<\/p>\n\n\n\n<ul><li><strong>Structure from motion\/\u57fa\u4e8e\u89c6\u9891\u7684\u6df1\u5ea6\u4f30\u8ba1<\/strong>\uff08\u65e0\u76d1\u7763\u5b66\u4e60\uff09<\/li><\/ul>\n\n\n\n<p>\u8fd9\u4e00\u90e8\u5206\u4e2d\u65e2\u5305\u542b\u4e86\u5355\u5e27\u89c6\u9891\u7684\u5355\u76ee\u6df1\u5ea6\u4f30\u8ba1\uff0c\u4e5f\u5305\u542b\u4e86\u591a\u5e27\u95f4\u89c6\u9891\u5e27\u7684\u50cf\u7d20\u7684\u7acb\u4f53\u5339\u914d\uff0c\u4ece\u800c\u8fd1\u4f3c\u83b7\u53d6\u591a\u89c6\u89d2\u56fe\u50cf\uff0c\u5bf9\u76f8\u673a\u4f4d\u59ff\u8fdb\u884c\u4f30\u8ba1\u3002<\/p>\n\n\n\n<div class=\"wp-block-image\"><figure class=\"aligncenter\"><img src=\"https:\/\/pic4.zhimg.com\/v2-e1f48a804aa793aa60ad595d4895b32b_r.jpg\" alt=\"\"\/><\/figure><\/div>\n\n\n\n<h4>\u8bc4\u4f30\u6307\u6807\uff1a<\/h4>\n\n\n\n<p>\u5728\u5355\u76ee\u6df1\u5ea6\u4f30\u8ba1\u95ee\u9898\u4e2d\uff0c\u5e38\u7528\u7684\u7cbe\u5ea6\u8bc4\u4f30\u6307\u6807\u6709\u76f8\u5bf9\u8bef\u5dee(REL)\u3001\u5747\u65b9\u6839\u8bef\u5dee(RMS)\u3001\u5bf9\u6570\u8bef\u5dee(LG)\u53ca\u9608\u503c\u8bef\u5dee(%&nbsp;correct)<\/p>\n\n\n\n<figure class=\"wp-block-image\"><img src=\"https:\/\/img-blog.csdnimg.cn\/20210901161932783.png?x-oss-process=image\/watermark,type_ZHJvaWRzYW5zZmFsbGJhY2s,shadow_50,text_Q1NETiBATWVuZ1lhX0RyZWFt,size_13,color_FFFFFF,t_70,g_se,x_16\" alt=\"\"\/><\/figure>\n\n\n\n<figure class=\"wp-block-image\"><img src=\"https:\/\/img-blog.csdnimg.cn\/20210901161953612.png?x-oss-process=image\/watermark,type_ZHJvaWRzYW5zZmFsbGJhY2s,shadow_50,text_Q1NETiBATWVuZ1lhX0RyZWFt,size_20,color_FFFFFF,t_70,g_se,x_16\" alt=\"\"\/><\/figure>\n\n\n\n<h3 id=\"%E6%B7%B1%E5%BA%A6%E4%BC%B0%E8%AE%A1%E7%9B%B8%E5%85%B3%E6%95%B0%E6%8D%AE%E9%9B%86%C2%A0\">\u6df1\u5ea6\u4f30\u8ba1\u76f8\u5173\u6570\u636e\u96c6&nbsp;<\/h3>\n\n\n\n<p>         \u5728\u6df1\u5ea6\u4f30\u8ba1\u7684\u7814\u7a76\u4e2d\uff0c\u7531\u4e8e\u5ba4\u5185\u5916\u573a\u666f\u7c7b\u578b\u4e0e\u6df1\u5ea6\u8303\u56f4\u5177\u6709\u8f83\u5927\u7684\u5dee\u5f02\uff0c\u5bf9\u5e94\u4e0d\u540c\u7684\u573a\u666f\u5206\u522b\u4f1a\u6784\u9020\u4e0d\u540c\u7684<strong>\u6570\u636e\u96c6<\/strong>\u3002<\/p>\n\n\n\n<ul><li><strong>\u771f\u5b9e\u573a\u666f\u6570\u636e\u96c6<\/strong><ul><li><strong>NYU depth v2<\/strong>\uff08\u6765\u81ea\u7ebd\u7ea6\u5927\u5b66\uff09\u662f\u5e38\u7528\u7684\u5ba4\u5185\u6570\u636e\u96c6\u4e4b\u4e00\uff0c<ul><li>\u9009\u53d6\u4e86464\u4e2a\u4e0d\u540c\u7684\u573a\u666f\uff0c<\/li><li>\u5229\u7528RGB\u76f8\u673a\u548c\u5fae\u8f6f\u7684Kinect\u6df1\u5ea6\u76f8\u673a\u540c\u65f6\u91c7\u96c6\u5ba4\u5185\u573a\u666f\u7684<strong>RGB\u4fe1\u606f<\/strong>\u548c<strong>\u6df1\u5ea6\u4fe1\u606f<\/strong>\uff0c\u6536\u96c6\u4e86407 024\u5e27<strong>RGBD<\/strong>\u56fe\u50cf\u5bf9\u6784\u5efa\u6570\u636e\u96c6\u3002<\/li><li>\u7531\u4e8e\u7ea2\u5916\u76f8\u673a\u548c\u6444\u50cf\u673a\u4e4b\u95f4\u7684\u4f4d\u7f6e\u504f\u5dee\uff0c\u6df1\u5ea6\u76f8\u673a\u91c7\u96c6\u7684\u539f\u59cb\u6df1\u5ea6\u56fe\u5b58\u5728\u7f3a\u5931\u90e8\u5206\u6216\u662f\u566a\u70b9\uff0c<\/li><li>\u4f5c\u8005\u4ece\u4e2d\u9009\u53d6\u4e861 449\u5e45\u56fe\u50cf\uff0c\u5229\u7528\u7740\u8272\u7b97\u6cd5\u5bf9\u6df1\u5ea6\u56fe\u8fdb\u884c\u586b\u5145\u5f97\u5230<strong>\u7a20\u5bc6\u6df1\u5ea6\u56fe<\/strong>\uff0c\u540c\u65f6<strong>\u4eba\u5de5\u6807\u6ce8\u8bed\u4e49\u4fe1\u606f<\/strong>\u3002<\/li><\/ul><\/li><li><strong>Make3D<\/strong>\uff08\u65af\u5766\u798f\u5927\u5b66\uff09\u662f\u5e38\u7528\u7684\u5ba4\u5916\u573a\u666f\u6570\u636e\u96c6\u4e4b\u4e00\uff0c<ul><li>\u4f7f\u7528\u6fc0\u5149\u626b\u63cf\u4eea\u91c7\u96c6\u5ba4\u5916\u573a\u666f\u7684\u6df1\u5ea6\u4fe1\u606f\uff0c<\/li><li>\u9009\u53d6\u7684\u573a\u666f\u7c7b\u578b\u4e3a\u767d\u5929\u7684\u57ce\u5e02\u548c\u81ea\u7136\u98ce\u5149\uff0c\u6df1\u5ea6\u8303\u56f4\u662f5~81 m\uff0c\u5927\u4e8e\u8be5\u8303\u56f4\u7edf\u4e00\u6620\u5c04\u4e3a81 m\u3002<\/li><li>\u6570\u636e\u96c6\u5171\u5305\u542b534\u5e45<strong>RGBD<\/strong>\u56fe\u50cf\u5bf9\uff0c\u5176\u4e2d400\u5e45\u7528\u4e8e\u8bad\u7ec3\uff0c134\u5e45\u7528\u4e8e\u6d4b\u8bd5\u3002<\/li><\/ul><\/li><li><strong>KITTI<\/strong>\uff08\u5fb7\u56fd\u5361\u5c14\u65af\u9c81\u5384\u7406\u5de5\u5b66\u9662\u548c\u7f8e\u56fd\u4e30\u7530\u6280\u672f\u7814\u7a76\u9662\uff09\u81ea\u52a8\u9a7e\u9a76\u9886\u57df\u5e38\u7528\u7684\u6570\u636e\u96c6\u4e4b\u4e00\uff0c\u94fe\u63a5\uff1a<a rel=\"noreferrer noopener\" href=\"http:\/\/www.cvlibs.net\/datasets\/kitti\/eval_object.php\" target=\"_blank\">http:\/\/www.cvlibs.net\/datasets\/kit<\/a><ul><li>\u5305\u542b\u6df1\u5ea6\u6570\u636e\u6807\u7b7e<\/li><li>\u901a\u8fc7\u4e00\u8f86\u88c5\u914d\u67092\u53f0\u9ad8\u5206\u8fa8\u7387\u5f69\u8272\u6444\u50cf\u673a\u30012\u53f0\u7070\u5ea6\u6444\u50cf\u673a\u3001\u6fc0\u5149\u626b\u63cf\u4eea\u548cGPS\u5b9a\u4f4d\u7cfb\u7edf\u7684\u6c7d\u8f66\u91c7\u96c6\u6570\u636e\uff0c\u5176\u4e2d\u6fc0\u5149\u626b\u63cf\u4eea\u7684\u6700\u5927\u6d4b\u91cf\u8ddd\u79bb\u4e3a120 m\u3002<\/li><li>\u56fe\u50cf\u573a\u666f\u5305\u62ec\u5361\u5c14\u65af\u9c81\u5384\u5e02\u3001\u91ce\u5916\u5730\u533a\u4ee5\u53ca\u9ad8\u901f\u516c\u8def\u3002<\/li><li>\u6570\u636e\u96c6\u5171\u5305\u542b93 000\u4e2a<strong>RGBD<\/strong>\u8bad\u7ec3\u6837\u672c\u3002<\/li><\/ul><\/li><li><strong>Depth in the Wild\uff08DIW\uff09<\/strong>\uff08\u5bc6\u6b47\u6839\u5927\u5b66\uff09\u4ee5<strong>\u76f8\u5bf9\u6df1\u5ea6<\/strong>\u4f5c\u4e3a\u6807\u7b7e\u7684\u6570\u636e\u96c6<ul><li>\u4ece\u8bcd\u5178\u4e2d\u968f\u673a\u9009\u53d6\u5355\u8bcd\u4f5c\u4e3a\u641c\u7d22\u5173\u952e\u5b57\uff0c\u7136\u540e\u4ece\u4e92\u8054\u7f51\u4e2d\u6536\u96c6\u5f97\u5230\u539f\u59cb\u7684RGB\u56fe\u50cf\u3002<\/li><li>\u6807\u6ce8\u5de5\u4f5c\u5916\u5305\u7ed9\u4e13\u95e8\u7684\u5de5\u4f5c\u4eba\u5458\uff0c\u4e3a\u4e86\u66f4\u52a0\u9ad8\u6548\uff0c\u6bcf\u4e00\u5e45\u56fe\u50cf\u9009\u53d6\u4e86\u4e24\u4e2a\u9ad8\u4eae\u7684\u70b9\uff0c\u5de5\u4f5c\u4eba\u5458\u53ea\u9700\u5224\u5b9a\u4e24\u4e2a\u70b9\u7684\u8fdc\u8fd1\u5173\u7cfb\u5373\u53ef\u3002<\/li><li>\u5bf9\u4e8e\u91c7\u6837\u70b9\u5bf9\u4e4b\u95f4\u7684\u4f4d\u7f6e\u5173\u7cfb\uff0c\u91c7\u752850%\u968f\u673a\u91c7\u6837\uff0c\u53e6\u591650%\u5bf9\u79f0\u91c7\u6837\u7684\u65b9\u6cd5\u4ee5\u4f7f\u6784\u5efa\u7684\u6570\u636e\u96c6\u5c3d\u53ef\u80fd\u5e73\u8861\u3002\u6700\u7ec8\u83b7\u5f97\u7684\u6709\u6548\u6807\u6ce8\u56fe\u50cf\u7ea65\u00d7E11\u5f20\u3002<\/li><\/ul><\/li><li><strong>Cityscapes<\/strong>. Cityscapes\u7684\u6570\u636e\u53d6\u81ea\u5fb7\u56fd\u768450\u591a\u4e2a\u57ce\u5e02\u7684\u6237\u5916\u573a\u666f\uff0c\u5176\u4e2d\u6570\u636e\u5305\u542b\u6709\u5de6\u53f3\u89c6\u89d2\u56fe\u50cf\u3001\u89c6\u5dee\u6df1\u5ea6\u56fe\u3001\u76f8\u673a\u6821\u51c6\u3001\u8f66\u8f86\u6d4b\u8ddd\u3001\u884c\u4eba\u6807\u5b9a\u3001\u76ee\u6807\u5206\u5272\u7b49\uff0c\u540c\u65f6\u4e5f\u5305\u542b\u6709\u7c7b\u4f3c\u4e8evKITTI\u7684\u865a\u62df\u6e32\u67d3\u573a\u666f\u56fe\u50cf\u3002\u5176\u4e2d\u7b80\u5355\u7684\u5de6\u89c6\u89d2\u56fe\u50cf\u3001\u76f8\u673a\u6807\u5b9a\u3001\u76ee\u6807\u5206\u5272\u7b49\u6570\u636e\u9700\u8981\u5229\u7528\u5b66\u751f\u8d26\u53f7\u6ce8\u518c\u83b7\u53d6\uff0c\u5176\u4ed6\u6570\u636e\u9700\u8981\u8054\u7cfb\u7ba1\u7406\u5458\u83b7\u53d6\u3002\u94fe\u63a5\uff1a<a rel=\"noreferrer noopener\" href=\"https:\/\/www.cityscapes-dataset.com\/downloads\/\" target=\"_blank\">https:\/\/www.cityscapes-dataset.com\/<\/a><\/li><\/ul><\/li><li><strong>\u865a\u62df\u573a\u666f\u6570\u636e\u96c6<\/strong><ul><li>SceneNet RGB-D\u6570\u636e\u96c6<\/li><li>SYNTHIA\u6570\u636e\u96c6<\/li><li>\u7531\u4e8e\u662f\u901a\u8fc7\u865a\u62df\u573a\u666f\u751f\u6210\uff0c\u6570\u636e\u96c6\u4e2d\u5305\u62ec\u66f4\u591a\u5929\u6c14\u3001\u73af\u5883\u53ca\u5149\u7167\uff0c\u573a\u666f\u7c7b\u578b\u591a\u6837\u3002\u5404\u6570\u636e\u96c6\u6709\u5404\u81ea\u7684\u4f18\u7f3a\u70b9\uff0c\u5728\u5b9e\u9645\u7814\u7a76\u4e2d\uff0c\u5e94\u6839\u636e\u5177\u4f53\u7814\u7a76\u95ee\u9898\u6765\u9009\u62e9\u5408\u9002\u7684\u6570\u636e\u96c6\u3002<\/li><\/ul><\/li><\/ul>\n\n\n\n<p>&nbsp;<strong>\u7efc\u4e0a\uff0c\u53ef\u4ee5\u770b\u5230\u57fa\u4e8e\u6df1\u5ea6\u5b66\u4e60\u7684\u5355\u76ee\u6df1\u5ea6\u4f30\u8ba1\u662f\u672c\u9886\u57df\u7684\u53d1\u5c55\u65b9\u5411\u3002\u76ee\u524d\uff0c\u8be5\u9886\u57df\u7684\u53d1\u5c55\u4e3b\u8981\u96c6\u4e2d\u5728\u6570\u636e\u96c6\u548c\u6df1\u5ea6\u5b66\u4e60\u6a21\u578b\u4e24\u65b9\u9762\u3002\u9996\u5148\uff0c\u6570\u636e\u96c6\u7684\u8d28\u91cf\u5728\u5f88\u5927\u7a0b\u5ea6\u4e0a\u51b3\u5b9a\u4e86\u6a21\u578b\u7684\u9c81\u68d2\u6027\u4e0e\u6cdb\u5316\u80fd\u529b\uff0c\u6df1\u5ea6\u5b66\u4e60\u8981\u6c42\u8bad\u7ec3\u6570\u636e\u5fc5\u987b\u6709\u66f4\u591a\u7684\u6570\u91cf\u3001\u66f4\u591a\u7684\u573a\u666f\u7c7b\u578b\uff0c\u5982\u4f55\u6784\u5efa\u6ee1\u8db3\u6df1\u5ea6\u5b66\u4e60\u7684\u6570\u636e\u96c6\u6210\u4e3a\u4e00\u4e2a\u91cd\u8981\u7684\u7814\u7a76\u65b9\u5411\u3002\u76ee\u524d\uff0c\u57fa\u4e8e\u865a\u62df\u573a\u666f\u751f\u6210\u6df1\u5ea6\u6570\u636e\u5177\u6709\u4e0d\u9700\u8981\u6602\u8d35\u7684\u6df1\u5ea6\u91c7\u96c6\u8bbe\u5907\u3001\u573a\u666f\u7c7b\u578b\u591a\u6837\u3001\u8282\u7701\u4eba\u529b\u6210\u672c\u7b49\u4f18\u52bf\uff0c\u7ed3\u5408\u771f\u5b9e\u573a\u666f\u548c\u865a\u62df\u573a\u666f\u7684\u6570\u636e\u5171\u540c\u8bad\u7ec3\u4e5f\u662f\u672a\u6765\u6df1\u5ea6\u5b66\u4e60\u65b9\u6cd5\u7684\u8d8b\u52bf\u3002\u5176\u6b21\uff0c\u4e3a\u4e86\u63d0\u9ad8\u6df1\u5ea6\u5b66\u4e60\u4f30\u8ba1\u5355\u5e45\u56fe\u50cf\u6df1\u5ea6\u7684\u7cbe\u5ea6\uff0c\u8981\u6c42\u66f4\u65b0\u7684\u66f4\u590d\u6742\u7684\u6df1\u5ea6\u6846\u67b6\u3002\u9664\u4e86\u795e\u7ecf\u7f51\u7edc\u6a21\u578b\u672c\u8eab\u7ed3\u6784\u7684\u4f18\u5316\uff0c\u66f4\u65b0\u9896\u7684\u7b97\u6cd5\u8bbe\u8ba1\u4e5f\u80fd\u6709\u6548\u5730\u63d0\u5347\u9884\u6d4b\u7cbe\u5ea6\u3002\u7814\u7a76\u5de5\u4f5c\u5927\u591a\u91c7\u7528\u6709\u76d1\u7763\u56de\u5f52\u6a21\u578b\u5bf9\u8fde\u7eed\u7684\u7edd\u5bf9\u6df1\u5ea6\u503c\u8fdb\u884c\u56de\u5f52\u62df\u5408\u3002\u8003\u8651\u5230\u573a\u666f\u7531\u8fdc\u53ca\u8fd1\u7684\u7279\u6027\uff0c\u4e5f\u6709\u7528\u5206\u7c7b\u6a21\u578b\u8fdb\u884c\u7edd\u5bf9\u6df1\u5ea6\u4f30\u8ba1\u7684\u65b9\u6cd5\u3002\u7531\u6df1\u5ea6\u4fe1\u606f\u548c\u5176\u4ed6\u4fe1\u606f\u4e4b\u95f4\u7684\u4e92\u8865\u6027\uff0c\u90e8\u5206\u5de5\u4f5c\u7ed3\u5408\u8868\u9762\u6cd5\u7ebf\u7b49\u4fe1\u606f\u63d0\u5347\u6df1\u5ea6\u9884\u6d4b\u7684\u7cbe\u5ea6\u3002\u6df1\u5ea6\u5b66\u4e60\u53d1\u5c55\u8fc5\u901f\uff0c\u65b0\u7684\u6a21\u578b\u5c42\u51fa\u4e0d\u7a77\uff0c\u5982\u4f55\u5c06\u8fd9\u4e9b\u6a21\u578b\u5e94\u7528\u4e8e\u5355\u5e45\u56fe\u50cf\u6df1\u5ea6\u4f30\u8ba1\u95ee\u9898\u4e2d\u9700\u8981\u66f4\u52a0\u6df1\u5165\u5730\u7814\u7a76\u3002\u53e6\u5916\uff0c\u63a2\u7d22\u795e\u7ecf\u7f51\u7edc\u5728\u5355\u76ee\u6df1\u5ea6\u4f30\u8ba1\u95ee\u9898\u4e2d\u5b66\u5230\u7684\u662f\u4f55\u79cd\u7279\u5f81\u4e5f\u662f\u4e00\u4e2a\u91cd\u8981\u7684\u7814\u7a76\u65b9\u5411\u3002<\/strong><\/p>\n\n\n\n<p><strong>\u5bf9\u4e8e\u5355\u76ee\u6df1\u5ea6\u4f30\u8ba1\u6a21\u578b\uff0c\u76ee\u524d\u4e3b\u8981\u5206\u4e3a\u57fa\u4e8e\u56de\u5f52\/\u5206\u7c7b\u7684\u76d1\u7763\u6a21\u578b\uff0c\u57fa\u4e8e\u53cc\u76ee\u8bad\u7ec3\/\u89c6\u9891\u5e8f\u5217\u7684\u65e0\u76d1\u7763\u6a21\u578b\uff0c\u4ee5\u53ca\u57fa\u4e8e\u751f\u6210\u5b66\u4e60\u7684\u56fe\u50cf\u98ce\u683c\u8fc1\u79fb\u6a21\u578b\u3002\u5927\u6982\u4ece2017\u5e74\u8d77\uff0c\u5373CVPR2018\u5f00\u59cb\uff0c\u5355\u76ee\u6df1\u5ea6\u4f30\u8ba1\u7684\u6548\u679c\u5c31\u5df2\u7ecf\u8fbe\u5230\u4e86\u53cc\u76ee\u6df1\u5ea6\u4f30\u8ba1\u7684\u6548\u679c\uff0c\u4e3b\u8981\u662f\u76d1\u7763\u6a21\u578b\u3002\u4f46\u662f\u7531\u4e8e\u73b0\u6709\u7684\u6570\u636e\u96c6\u4e3b\u8981\u4e3aKITTI\u3001Cityscapes\u3001NYU DepthV2\u7b49\uff0c\u5176\u573a\u666f\u548c\u76f8\u673a\u90fd\u662f\u56fa\u5b9a\u7684\uff0c\u4ece\u800c\u5bfc\u81f4\u76d1\u7763\u5b66\u4e60\u4e0b\u7684\u6a21\u578b\u65e0\u6cd5\u9002\u7528\u4e8e\u5176\u4ed6\u573a\u666f\uff0c\u5c24\u5176\u662f\u591a\u76ee\u6807\u8ddf\u8e2a\u8fd9\u7c7b\u7ec6\u8282\u4e30\u5bcc\u7684\u573a\u666f\uff0c\u53ef\u4ee5\u4ece\u8bba\u6587\u4e2d\u770b\u5230\uff0c\u57fa\u672c\u4e0a\u6bcf\u4e2a\u6570\u636e\u96c6\u90fd\u4f1a\u6709\u4e00\u4e2a\u5355\u72ec\u7684\u9884\u8bad\u7ec3\u6a21\u578b\u3002<\/strong><\/p>\n\n\n\n<p>\u5bf9\u4e8eGAN\uff0c\u5176\u5bf9\u4e8e\u56fe\u50cf\u98ce\u683c\u7684\u8fc1\u79fb\u672c\u8eab\u662f\u4e00\u4e2a\u5f88\u597d\u7684\u6cdb\u5316\u70b9\uff0c\u65e2\u53ef\u4ee5\u7528\u4e8e\u5c06\u573a\u666f\u53d8\u4e3a\u6674\u5929\u3001\u96fe\u5929\u7b49\u60c5\u51b5\uff0c\u4e5f\u53ef\u4ee5\u7528\u4e8e\u56fe\u50cf\u5206\u5272\u573a\u666f\u3002\u4f46\u662f\u6df1\u5ea6\u4f30\u8ba1\u95ee\u9898\u4e2d\uff0c\u50cf\u7d20\u70b9\u5b58\u5728\u76f8\u5bf9\u5927\u5c0f\uff0c\u56e0\u6b64\u5fc5\u5b9a\u6d89\u53ca\u5230\u56de\u5f52\uff0c\u56e0\u6b64\u5176\u5fc5\u5b9a\u662f\u76d1\u7763\u5b66\u4e60\u6a21\u578b\uff0c\u6240\u4ee5\u6cdb\u5316\u6027\u80fd\u4e5f\u4e0d\u597d\u3002\u5bf9\u4e8e\u65e0\u76d1\u7763\u7684\u7b97\u6cd5\uff0c\u53ef\u80fd\u573a\u666f\u9002\u5e94\u6027\u4f1a\u66f4\u597d\uff0c\u4f46\u4f9d\u65e7\u4e0d\u9002\u7528\u4e8e\u5bf9\u884c\u4eba\u6df1\u5ea6\u7684\u4f30\u8ba1\u3002<\/p>\n\n\n\n<h4><strong>\u53c2\u8003\u6587\u732e<\/strong><\/h4>\n\n\n\n<p>[1] Long J, Shelhamer E, Darrell T. Fully convolutional networks for semantic segmentation[C]\/\/Proceedings of the IEEE conference on computer vision and pattern recognition. 2015: 3431-3440.<\/p>\n\n\n\n<p>[2] Ronneberger O, Fischer P, Brox T. U-net: Convolutional networks for biomedical image segmentation[C]\/\/International Conference on Medical image computing and computer-assisted intervention. Springer, Cham, 2015: 234-241.<\/p>\n\n\n\n<p>[3] Laina I, Rupprecht C, Belagiannis V, et al. Deeper depth prediction with fully convolutional residual networks[C]\/\/2016 Fourth international conference on 3D vision (3DV). IEEE, 2016: 239-248.<\/p>\n\n\n\n<p>[4] Fu H, Gong M, Wang C, et al. Deep ordinal regression network for monocular depth estimation[C]\/\/Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. 2018: 2002-2011.<\/p>\n\n\n\n<p>[5] Godard C, Mac Aodha O, Brostow G J. Unsupervised monocular depth estimation with left-right consistency[C]\/\/Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. 2017: 270-279.<\/p>\n\n\n\n<p>[6] Dosovitskiy A, Fischer P, Ilg E, et al. Flownet: Learning optical flow with convolutional networks[C]\/\/Proceedings of the IEEE international conference on computer vision. 2015: 2758-2766.<\/p>\n\n\n\n<p>[7] Ilg E, Mayer N, Saikia T, et al. Flownet 2.0: Evolution of optical flow estimation with deep networks[C]\/\/Proceedings of the IEEE conference on computer vision and pattern recognition. 2017: 2462-2470.<\/p>\n\n\n\n<p>[8] Mayer N, Ilg E, Hausser P, et al. A large dataset to train convolutional networks for disparity, optical flow, and scene flow estimation[C]\/\/Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. 2016: 4040-4048.<\/p>\n\n\n\n<p>[9] Xie J, Girshick R, Farhadi A. Deep3d: Fully automatic 2d-to-3d video conversion with deep convolutional neural networks[C]\/\/European Conference on Computer Vision. Springer, Cham, 2016: 842-857.<\/p>\n\n\n\n<p>[10] Luo Y, Ren J, Lin M, et al. Single View Stereo Matching[C]\/\/Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. 2018.<\/p>\n\n\n\n<p>[11] Zhou T, Brown M, Snavely N, et al. Unsupervised learning of depth and ego-motion from video[C]\/\/Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. 2017: 1851-1858.<\/p>\n\n\n\n<p>[12] Yin Z, Shi J. Geonet: Unsupervised learning of dense depth, optical flow and camera pose[C]\/\/Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. 2018: 1983-1992.<\/p>\n\n\n\n<p>[13] Zhan H, Garg R, Saroj Weerasekera C, et al. Unsupervised learning of monocular depth estimation and visual odometry with deep feature reconstruction[C]\/\/ Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. 2018: 340-349.<\/p>\n\n\n\n<p>[14] Goodfellow I, Pouget-Abadie J, Mirza M, et al. Generative adversarial nets[C]\/\/Advances in neural information processing systems. 2014: 2672-2680.<\/p>\n\n\n\n<p>[15] Radford A , Metz L , Chintala S . Unsupervised Representation Learning with Deep Convolutional Generative Adversarial Networks[J]. Computer Science, 2015.<\/p>\n\n\n\n<p>[16] Arjovsky M, Chintala S, Bottou L. Wasserstein gan[J]. arXiv preprint arXiv:1701.07875, 2017.<\/p>\n\n\n\n<p>[17] Gulrajani I, Ahmed F, Arjovsky M, et al. Improved training of wasserstein gans[C]\/\/Advances in Neural Information Processing Systems. 2017: 5767-5777.<\/p>\n\n\n\n<p>[18] Mao X, Li Q, Xie H, et al. Least squares generative adversarial networks[C]\/\/Proceedings of the IEEE International Conference on Computer Vision. 2017: 2794-2802.<\/p>\n\n\n\n<p>[19] Mirza M, Osindero S. Conditional generative adversarial nets[J]. arXiv preprint arXiv:1411.1784, 2014.<\/p>\n\n\n\n<p>[20] Isola P, Zhu J Y, Zhou T, et al. Image-to-image translation with conditional adversarial networks[C]\/\/Proceedings of the IEEE conference on computer vision and pattern recognition. 2017: 1125-1134.<\/p>\n\n\n\n<p>[21] Wang T C, Liu M Y, Zhu J Y, et al. High-resolution image synthesis and semantic manipulation with conditional gans[C]\/\/Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. 2018: 8798-8807.<\/p>\n\n\n\n<p>[22] Zhu J Y, Park T, Isola P, et al. Unpaired image-to-image translation using cycle-consistent adversarial networks[C]\/\/Proceedings of the IEEE international conference on computer vision. 2017: 2223-2232.<\/p>\n\n\n\n<p>[23] Wang T C , Liu M Y , Zhu J Y , et al. Video-to-Video Synthesis[J]. arXiv preprint arXiv:1808.06601,2018.<\/p>\n\n\n\n<p>[24] Zheng C, Cham T J, Cai J. T2net: Synthetic-to-realistic translation for solving single-image depth estimation tasks[C]\/\/Proceedings of the European Conference on Computer Vision (ECCV). 2018: 767-783.<\/p>\n\n\n\n<p>[25] Atapour-Abarghouei A, Breckon T P. Real-time monocular depth estimation using synthetic data with domain adaptation via image style transfer[C]\/\/Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. 2018: 2800-2810.<\/p>\n\n\n\n<p>[26] Nekrasov V , Dharmasiri T , Spek A , et al. Real-Time Joint Semantic Segmentation and Depth Estimation Using Asymmetric Annotations[J]. arXiv preprint arXiv:1809.04766,2018.<\/p>\n\n\n\n<p>[27] Nekrasov V , Shen C , Reid I . Light-Weight RefineNet for Real-Time Semantic Segmentation[J]. arXiv preprint arXiv:1810.03272, 2018.<\/p>\n\n\n\n<p>[28] Lin G , Milan A , Shen C , et al. RefineNet: Multi-Path Refinement Networks for High-Resolution Semantic Segmentation[C]\/\/Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition.,2017:1925-1934<\/p>\n\n\n\n<p>[29] Zou Y , Luo Z , Huang J B . DF-Net: Unsupervised Joint Learning of Depth and Flow using Cross-Task Consistency[C]\/\/Proceedings of the European Conference on Computer Vision (ECCV). 2018:36-53.<\/p>\n\n\n\n<p>[30] Ranjan A, Jampani V, Balles L, et al. Competitive collaboration: Joint unsupervised learning of depth, camera motion, optical flow and motion segmentation[C]\/\/Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. 2019: 12240-12249.<\/p>\n","protected":false},"excerpt":{"rendered":"<p>Monocular Depth Estimation Monocular Depth Estimation i &hellip; <a href=\"http:\/\/139.9.1.231\/index.php\/2023\/01\/30\/monocular-depth-estimation\/\" class=\"more-link\">\u7ee7\u7eed\u9605\u8bfb<span class=\"screen-reader-text\">\u57fa\u4e8e\u6df1\u5ea6\u5b66\u4e60\u7684\u5355\u76ee\u6df1\u5ea6\u4f30\u8ba1\u7efc\u8ff0<\/span><\/a><\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":[],"categories":[4,36,9],"tags":[],"_links":{"self":[{"href":"http:\/\/139.9.1.231\/index.php\/wp-json\/wp\/v2\/posts\/12171"}],"collection":[{"href":"http:\/\/139.9.1.231\/index.php\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"http:\/\/139.9.1.231\/index.php\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"http:\/\/139.9.1.231\/index.php\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"http:\/\/139.9.1.231\/index.php\/wp-json\/wp\/v2\/comments?post=12171"}],"version-history":[{"count":39,"href":"http:\/\/139.9.1.231\/index.php\/wp-json\/wp\/v2\/posts\/12171\/revisions"}],"predecessor-version":[{"id":12246,"href":"http:\/\/139.9.1.231\/index.php\/wp-json\/wp\/v2\/posts\/12171\/revisions\/12246"}],"wp:attachment":[{"href":"http:\/\/139.9.1.231\/index.php\/wp-json\/wp\/v2\/media?parent=12171"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"http:\/\/139.9.1.231\/index.php\/wp-json\/wp\/v2\/categories?post=12171"},{"taxonomy":"post_tag","embeddable":true,"href":"http:\/\/139.9.1.231\/index.php\/wp-json\/wp\/v2\/tags?post=12171"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}