{"id":785,"date":"2020-10-10T10:29:00","date_gmt":"2020-10-10T01:29:00","guid":{"rendered":"https:\/\/arithmer.blog\/?p=785"},"modified":"2022-03-08T15:45:10","modified_gmt":"2022-03-08T06:45:10","slug":"object-pose-estimation","status":"publish","type":"post","link":"https:\/\/arithmer.blog\/blog\/object-pose-estimation","title":{"rendered":"\u525b\u4f53\u306e\u59ff\u52e2\u63a8\u5b9a"},"content":{"rendered":"\n<p class=\"has-small-font-size\">\u672c\u8cc7\u6599\u306f2020\u5e7410\u670810\u65e5\u306b\u793e\u5185\u5171\u6709\u8cc7\u6599\u3068\u3057\u3066\u5c55\u958b\u3057\u3066\u3044\u305f\u3082\u306e\u3092WEB\u30da\u30fc\u30b8\u5411\u3051\u306b\u30ea\u30cb\u30e5\u30fc\u30a2\u30eb\u3057\u305f\u5185\u5bb9\u306b\u306a\u308a\u307e\u3059\u3002<\/p>\n\n\n\n<h3 class=\"has-medium-font-size wp-block-heading\" id=\"object-pose-estimation\"><strong>\u25a0Object Pose Estimation<\/strong><\/h3>\n\n\n\n<figure class=\"wp-block-image size-full\"><img decoding=\"async\" width=\"1024\" height=\"220\" src=\"https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20201010_01.jpg\" alt=\"\" class=\"wp-image-789\" srcset=\"https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20201010_01.jpg 1024w, https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20201010_01-300x64.jpg 300w, https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20201010_01-768x165.jpg 768w, https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20201010_01-304x65.jpg 304w\" sizes=\"(max-width: 1024px) 100vw, 1024px\"><\/figure>\n\n\n\n<ul style=\"font-size:16px\"><li>Correspondence-based method : Find Correspondences between 2D points and 3D points (PnP, ICP)<\/li><li>Template-based method: Extract the gradient information for matching, and refined(ICP)<\/li><li>Voting-based method: Every local predicts a result, and refined using RANSAC<\/li><li>Regression-based method: Represent pose suitable for CNN<\/li><\/ul>\n\n\n\n<h3 class=\"has-medium-font-size wp-block-heading\" id=\"summary-of-6d-pose-estimation-methods\"><strong>\u25a0Summary of 6D pose estimation methods<\/strong><\/h3>\n\n\n\n<figure class=\"wp-block-image size-full\"><img decoding=\"async\" width=\"1024\" height=\"427\" src=\"https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20201010_02.jpg\" alt=\"\" class=\"wp-image-790\" srcset=\"https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20201010_02.jpg 1024w, https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20201010_02-300x125.jpg 300w, https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20201010_02-768x320.jpg 768w, https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20201010_02-304x127.jpg 304w\" sizes=\"(max-width: 1024px) 100vw, 1024px\"><\/figure>\n\n\n\n<p class=\"has-text-align-center has-small-font-size\">Arxiv:1905.06658<\/p>\n\n\n\n<h3 class=\"has-medium-font-size wp-block-heading\" id=\"correspondence-based-method\"><strong>\u25a0Correspondence-based method<\/strong><\/h3>\n\n\n\n<p style=\"font-size:16px\">With rich textures to get image features<\/p>\n\n\n\n<figure class=\"wp-block-image size-full\"><img decoding=\"async\" width=\"1024\" height=\"379\" src=\"https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20201010_03.jpg\" alt=\"\" class=\"wp-image-791\" srcset=\"https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20201010_03.jpg 1024w, https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20201010_03-300x111.jpg 300w, https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20201010_03-768x284.jpg 768w, https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20201010_03-304x113.jpg 304w\" sizes=\"(max-width: 1024px) 100vw, 1024px\"><\/figure>\n\n\n\n<p class=\"has-text-align-center has-small-font-size\">Arxiv:1905.06658<\/p>\n\n\n\n<h3 class=\"has-medium-font-size wp-block-heading\" id=\"pnp-solver\"><strong>\u25a0PnP solver<\/strong><\/h3>\n\n\n\n<p style=\"font-size:16px\">Estimate extrinsic camera parameter using known 3D points in a world coordinate<br><br>There are many many solvers (e.g. P3P, P5P.., using RANSAC or Eigen value decomposition)<\/p>\n\n\n\n<figure class=\"wp-block-image size-full\"><img decoding=\"async\" width=\"1024\" height=\"291\" src=\"https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20201010_04.jpg\" alt=\"\" class=\"wp-image-792\" srcset=\"https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20201010_04.jpg 1024w, https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20201010_04-300x85.jpg 300w, https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20201010_04-768x218.jpg 768w, https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20201010_04-304x86.jpg 304w\" sizes=\"(max-width: 1024px) 100vw, 1024px\"><\/figure>\n\n\n\n<p class=\"has-text-align-center has-small-font-size\">Arxiv:1905.06658<\/p>\n\n\n\n<h3 class=\"has-medium-font-size wp-block-heading\" id=\"icp-algorithm\"><strong>\u25a0ICP algorithm<\/strong><\/h3>\n\n\n\n<p style=\"font-size:16px\">Iterative matching for 2 point clouds<\/p>\n\n\n\n<ol style=\"font-size:16px\"><li>Correspondence nearest neighbor points<\/li><li>Translate a point cloud set to match the other sets<\/li><li>iteration above<\/li><\/ol>\n\n\n\n<figure class=\"wp-block-image size-full\"><img decoding=\"async\" width=\"1024\" height=\"294\" src=\"https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20201010_05.jpg\" alt=\"\" class=\"wp-image-793\" srcset=\"https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20201010_05.jpg 1024w, https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20201010_05-300x86.jpg 300w, https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20201010_05-768x221.jpg 768w, https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20201010_05-304x87.jpg 304w\" sizes=\"(max-width: 1024px) 100vw, 1024px\"><\/figure>\n\n\n\n<h3 class=\"has-medium-font-size wp-block-heading\" id=\"icp\"><strong>\u25a0ICP<\/strong><\/h3>\n\n\n\n<p style=\"font-size:16px\">Estimate extrinsic camera parameter using known 3D points in a world coordinate There are many many solvers (e.g. P3P, P5P.., using RANSAC or Eigen value decomposition)<\/p>\n\n\n\n<p><\/p>\n\n\n\n<figure class=\"wp-block-image size-full\"><img decoding=\"async\" width=\"1024\" height=\"290\" src=\"https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20201010_06.jpg\" alt=\"\" class=\"wp-image-794\" srcset=\"https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20201010_06.jpg 1024w, https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20201010_06-300x85.jpg 300w, https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20201010_06-768x218.jpg 768w, https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20201010_06-304x86.jpg 304w\" sizes=\"(max-width: 1024px) 100vw, 1024px\"><\/figure>\n\n\n\n<p class=\"has-text-align-center has-small-font-size\">Arxiv:1905.06658<\/p>\n\n\n\n<h3 class=\"has-medium-font-size wp-block-heading\" id=\"segicp-segnet\"><strong>\u25a0SegICP\u3000SegNet<\/strong><\/h3>\n\n\n\n<p style=\"font-size:16px\"><strong>SegICP<\/strong><\/p>\n\n\n\n<p style=\"font-size:16px\">\u3000\u3000Segmentation and depth image-based partial Registration to obtain object\u2019s pose<\/p>\n\n\n\n<p style=\"font-size:16px\"><strong>SegNet<\/strong><br>\u3000\u3000Multi-hypothesis object pose: Point-to-point ICP<br><br>\u3000\u30001cm pose error and &lt; 5\u00b0error<\/p>\n\n\n\n<figure class=\"wp-block-image size-full\"><img decoding=\"async\" width=\"1024\" height=\"358\" src=\"https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20201010_07.jpg\" alt=\"\" class=\"wp-image-795\" srcset=\"https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20201010_07.jpg 1024w, https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20201010_07-300x105.jpg 300w, https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20201010_07-768x269.jpg 768w, https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20201010_07-304x106.jpg 304w\" sizes=\"(max-width: 1024px) 100vw, 1024px\"><\/figure>\n\n\n\n<h3 class=\"has-medium-font-size wp-block-heading\" id=\"template-based-method\"><strong>\u25a0Template-based method<\/strong><\/h3>\n\n\n\n<p style=\"font-size:16px\">Using gradient information in RGB-D or RGB image Without rich texture<\/p>\n\n\n\n<figure class=\"wp-block-image size-full\"><img decoding=\"async\" width=\"1024\" height=\"281\" src=\"https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20201010_08.jpg\" alt=\"\" class=\"wp-image-796\" srcset=\"https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20201010_08.jpg 1024w, https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20201010_08-300x82.jpg 300w, https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20201010_08-768x211.jpg 768w, https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20201010_08-304x83.jpg 304w\" sizes=\"(max-width: 1024px) 100vw, 1024px\"><\/figure>\n\n\n\n<p style=\"font-size:18px\"><strong>ICP<\/strong><\/p>\n\n\n\n<figure class=\"wp-block-image size-full\"><img decoding=\"async\" width=\"1024\" height=\"227\" src=\"https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20201010_09.jpg\" alt=\"\" class=\"wp-image-797\" srcset=\"https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20201010_09.jpg 1024w, https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20201010_09-300x67.jpg 300w, https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20201010_09-768x170.jpg 768w, https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20201010_09-304x67.jpg 304w\" sizes=\"(max-width: 1024px) 100vw, 1024px\"><\/figure>\n\n\n\n<p class=\"has-text-align-center has-small-font-size\">Arxiv:1905.06658<\/p>\n\n\n\n<h3 class=\"has-medium-font-size wp-block-heading\" id=\"voting-based-method\"><strong>\u25a0Voting-based method<\/strong><br><\/h3>\n\n\n\n<p style=\"font-size:16px\">Objects with occlusions Dense class labeling<\/p>\n\n\n\n<figure class=\"wp-block-image size-full\"><img decoding=\"async\" width=\"1024\" height=\"275\" src=\"https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20201010_10.jpg\" alt=\"\" class=\"wp-image-798\" srcset=\"https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20201010_10.jpg 1024w, https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20201010_10-300x81.jpg 300w, https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20201010_10-768x206.jpg 768w, https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20201010_10-304x82.jpg 304w\" sizes=\"(max-width: 1024px) 100vw, 1024px\"><\/figure>\n\n\n\n<p class=\"has-text-align-center has-small-font-size\">Arxiv:1905.06658<\/p>\n\n\n\n<h3 class=\"has-medium-font-size wp-block-heading\" id=\"regression-based-method\"><strong>\u25a0Regression-based method<\/strong><\/h3>\n\n\n\n<p style=\"font-size:16px\">PoseCNN<\/p>\n\n\n\n<ul style=\"font-size:16px\"><li>Localizing the center of objects<\/li><li>Predicting its distance from camera<\/li><li>Regressing rotation to a quarternion<\/li><\/ul>\n\n\n\n<figure class=\"wp-block-image size-full\"><img decoding=\"async\" width=\"1024\" height=\"286\" src=\"https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20201010_11.jpg\" alt=\"\" class=\"wp-image-799\" srcset=\"https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20201010_11.jpg 1024w, https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20201010_11-300x84.jpg 300w, https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20201010_11-768x215.jpg 768w, https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20201010_11-304x85.jpg 304w\" sizes=\"(max-width: 1024px) 100vw, 1024px\"><\/figure>\n\n\n\n<p style=\"font-size:16px\">PoseCNN architecture<\/p>\n\n\n\n<ul style=\"font-size:16px\"><li>Feature extraction<\/li><li>Embedding<\/li><li>Classification and regression<\/li><\/ul>\n\n\n\n<p>Two loss function for quaternion<\/p>\n\n\n\n<ul style=\"font-size:16px\"><li>PoseLoss<\/li><li>ShapeMatch-Loss<\/li><\/ul>\n\n\n\n<figure class=\"wp-block-image size-full\"><img decoding=\"async\" width=\"1024\" height=\"351\" src=\"https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20201010_12.jpg\" alt=\"\" class=\"wp-image-800\" srcset=\"https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20201010_12.jpg 1024w, https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20201010_12-300x103.jpg 300w, https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20201010_12-768x263.jpg 768w, https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20201010_12-304x104.jpg 304w\" sizes=\"(max-width: 1024px) 100vw, 1024px\"><\/figure>\n\n\n\n<p style=\"font-size:16px\">DeepIM<br>Iterative refinement using DNN<br>Enable to combine other estimator<br><\/p>\n\n\n\n<figure class=\"wp-block-image size-full\"><img decoding=\"async\" width=\"1024\" height=\"354\" src=\"https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20201010_13.jpg\" alt=\"\" class=\"wp-image-801\" srcset=\"https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20201010_13.jpg 1024w, https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20201010_13-300x104.jpg 300w, https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20201010_13-768x266.jpg 768w, https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20201010_13-304x105.jpg 304w\" sizes=\"(max-width: 1024px) 100vw, 1024px\"><\/figure>\n\n\n\n<p style=\"font-size:16px\">DeepIM <br>Iterative refinement using DNN <br><br>Flownet: estimate optical flow<\/p>\n\n\n\n<figure class=\"wp-block-image size-full\"><img decoding=\"async\" width=\"1024\" height=\"391\" src=\"https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20201010_14.jpg\" alt=\"\" class=\"wp-image-802\" srcset=\"https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20201010_14.jpg 1024w, https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20201010_14-300x115.jpg 300w, https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20201010_14-768x293.jpg 768w, https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20201010_14-304x116.jpg 304w\" sizes=\"(max-width: 1024px) 100vw, 1024px\"><\/figure>\n\n\n\n<p>CullNet<br>Single view image-based object pose estimation <br>Culling false positives Calibrate the confidence score<\/p>\n\n\n\n<figure class=\"wp-block-image size-full\"><img decoding=\"async\" width=\"1024\" height=\"263\" src=\"https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20201010_15.jpg\" alt=\"\" class=\"wp-image-803\" srcset=\"https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20201010_15.jpg 1024w, https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20201010_15-300x77.jpg 300w, https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20201010_15-768x197.jpg 768w, https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20201010_15-304x78.jpg 304w\" sizes=\"(max-width: 1024px) 100vw, 1024px\"><\/figure>\n\n\n\n<h3 class=\"has-medium-font-size wp-block-heading\" id=\"correspondence-based-and-regression-based-method\"><strong>\u25a0Correspondence-based and Regression-based method<\/strong><\/h3>\n\n\n\n<figure class=\"wp-block-image size-full\"><img decoding=\"async\" width=\"1024\" height=\"333\" src=\"https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20201010_16.jpg\" alt=\"\" class=\"wp-image-804\" srcset=\"https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20201010_16.jpg 1024w, https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20201010_16-300x98.jpg 300w, https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20201010_16-768x250.jpg 768w, https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20201010_16-304x99.jpg 304w\" sizes=\"(max-width: 1024px) 100vw, 1024px\"><\/figure>\n\n\n\n<p id=\"cullnet-architectureusing-yolov3-for-keypointscullnet-output-confidence-of-objects-pose-by-pnp-voting-based-and-regression-based-method\">CullNet Architecture<br><br>Using Yolov3 for keypoints<br><br>Cullnet output confidence of objects pose by PnP.<\/p>\n\n\n\n<p id=\"cullnet-architectureusing-yolov3-for-keypointscullnet-output-confidence-of-objects-pose-by-pnp-voting-based-and-regression-based-method\"><\/p>\n\n\n\n<h3 class=\"has-medium-font-size wp-block-heading\" id=\"cullnet-architectureusing-yolov3-for-keypointscullnet-output-confidence-of-objects-pose-by-pnp-voting-based-and-regression-based-method\"><strong>\u25a0Voting-based and Regression-based method<\/strong><\/h3>\n\n\n\n<figure class=\"wp-block-image size-full\"><img decoding=\"async\" width=\"1024\" height=\"320\" src=\"https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20201010_17.jpg\" alt=\"\" class=\"wp-image-805\" srcset=\"https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20201010_17.jpg 1024w, https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20201010_17-300x94.jpg 300w, https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20201010_17-768x240.jpg 768w, https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20201010_17-304x95.jpg 304w\" sizes=\"(max-width: 1024px) 100vw, 1024px\"><\/figure>\n\n\n\n<p style=\"font-size:16px\">DenseFusion<br>Estimating 6D pose of known objects<\/p>\n\n\n\n<h3 class=\"has-medium-font-size wp-block-heading\" id=\"comparison\"><strong>\u25a0Comparison<\/strong><\/h3>\n\n\n\n<p style=\"font-size:16px\">ADD (Average Distance of Model Point)<br>ADD-S (for symmetric objects<br><br>Error smaller than threshold(&lt; 2cm)<\/p>\n\n\n\n<figure class=\"wp-block-image size-full\"><img decoding=\"async\" width=\"1022\" height=\"53\" src=\"https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20201010_18.jpg\" alt=\"\" class=\"wp-image-806\" srcset=\"https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20201010_18.jpg 1022w, https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20201010_18-300x16.jpg 300w, https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20201010_18-768x40.jpg 768w, https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20201010_18-304x16.jpg 304w\" sizes=\"(max-width: 1022px) 100vw, 1022px\"><\/figure>\n\n\n\n<p style=\"font-size:16px\">From the tables, we can see that DenseFusion achieved the highest accuracy comparing with other regression-based methods. novel local feature fusion scheme using both RGB and depth images<\/p>\n\n\n\n<figure class=\"wp-block-image size-full\"><img decoding=\"async\" width=\"1024\" height=\"287\" src=\"https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20201010_19.jpg\" alt=\"\" class=\"wp-image-807\" srcset=\"https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20201010_19.jpg 1024w, https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20201010_19-300x84.jpg 300w, https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20201010_19-768x215.jpg 768w, https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20201010_19-304x85.jpg 304w\" sizes=\"(max-width: 1024px) 100vw, 1024px\"><\/figure>\n\n\n\n<h3 class=\"has-medium-font-size wp-block-heading\" id=\"\u30c0\u30a6\u30f3\u30ed\u30fc\u30c9\"><strong>\u25a0\u30c0\u30a6\u30f3\u30ed\u30fc\u30c9<\/strong><\/h3>\n\n\n\n<p><a href=\"https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/12_\u525b\u4f53\u306e\u59ff\u52e2\u63a8\u5b9a.pdf\" target=\"_blank\" rel=\"noreferrer noopener\">\u525b\u4f53\u306e\u59ff\u52e2\u63a8\u5b9a.pdf<\/a><\/p>\n","protected":false},"excerpt":{"rendered":"<p>\u672c\u8cc7\u6599\u306f2020\u5e7410\u670810\u65e5\u306b\u793e\u5185\u5171\u6709\u8cc7\u6599\u3068\u3057\u3066\u5c55\u958b\u3057\u3066\u3044\u305f\u3082\u306e\u3092WEB\u30da\u30fc\u30b8\u5411\u3051\u306b\u30ea\u30cb\u30e5\u30fc\u30a2\u30eb\u3057\u305f\u5185\u5bb9\u306b\u306a\u308a\u307e\u3059\u3002 \u25a0Object Pose Estimation Correspondence-based metho &#8230; <\/p>\n","protected":false},"author":3,"featured_media":788,"comment_status":"closed","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[13],"tags":[20,66,67,35,37,24,36],"_links":{"self":[{"href":"https:\/\/arithmer.blog\/index.php?rest_route=\/wp\/v2\/posts\/785"}],"collection":[{"href":"https:\/\/arithmer.blog\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/arithmer.blog\/index.php?rest_route=\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/arithmer.blog\/index.php?rest_route=\/wp\/v2\/users\/3"}],"replies":[{"embeddable":true,"href":"https:\/\/arithmer.blog\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=785"}],"version-history":[{"count":5,"href":"https:\/\/arithmer.blog\/index.php?rest_route=\/wp\/v2\/posts\/785\/revisions"}],"predecessor-version":[{"id":815,"href":"https:\/\/arithmer.blog\/index.php?rest_route=\/wp\/v2\/posts\/785\/revisions\/815"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/arithmer.blog\/index.php?rest_route=\/wp\/v2\/media\/788"}],"wp:attachment":[{"href":"https:\/\/arithmer.blog\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=785"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/arithmer.blog\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=785"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/arithmer.blog\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=785"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}