{"id":78,"date":"2021-12-09T14:17:25","date_gmt":"2021-12-09T14:17:25","guid":{"rendered":"https:\/\/mscvprojects.ri.cmu.edu\/2021teame\/?page_id=78"},"modified":"2022-04-29T19:30:28","modified_gmt":"2022-04-29T19:30:28","slug":"fall-2021","status":"publish","type":"page","link":"https:\/\/mscvprojects.ri.cmu.edu\/2021teame\/fall-2021\/","title":{"rendered":"Fall 2021"},"content":{"rendered":"\n<p><strong><span class=\"has-inline-color has-vivid-cyan-blue-color\">Summary<\/span><\/strong><\/p>\n\n\n\n<p>In this semester, our progress can be categorized into two folds: <br>(1) Panoptic Segmentation <br>(2) From Panoptic Segmentor to Detector<\/p>\n\n\n\n<p>By the end of this semester, we also submitted our ongoing version of Panoptic Segmentor to <strong><a href=\"https:\/\/driving-olympics.ai\/\">NeurIPS &#8217;21 AI Driving Olympics Workshop<\/a><\/strong>, and ranked <strong><span class=\"has-inline-color has-vivid-red-color\">TOP 3<\/span><\/strong> in both LiDAR-track and Open-track (we are <em>Team_AX_Semantic<\/em>).<\/p>\n\n\n\n<figure class=\"wp-block-image\"><img loading=\"lazy\" decoding=\"async\" width=\"1024\" height=\"330\" src=\"https:\/\/mscvprojects.ri.cmu.edu\/2021teame\/wp-content\/uploads\/sites\/50\/2021\/12\/Screen-Shot-2021-12-10-at-5.19.37-PM-1024x330.png\" alt=\"\" class=\"wp-image-112\" srcset=\"https:\/\/mscvprojects.ri.cmu.edu\/2021teame\/wp-content\/uploads\/sites\/50\/2021\/12\/Screen-Shot-2021-12-10-at-5.19.37-PM-1024x330.png 1024w, https:\/\/mscvprojects.ri.cmu.edu\/2021teame\/wp-content\/uploads\/sites\/50\/2021\/12\/Screen-Shot-2021-12-10-at-5.19.37-PM-300x97.png 300w, https:\/\/mscvprojects.ri.cmu.edu\/2021teame\/wp-content\/uploads\/sites\/50\/2021\/12\/Screen-Shot-2021-12-10-at-5.19.37-PM-768x248.png 768w, https:\/\/mscvprojects.ri.cmu.edu\/2021teame\/wp-content\/uploads\/sites\/50\/2021\/12\/Screen-Shot-2021-12-10-at-5.19.37-PM-1536x496.png 1536w, https:\/\/mscvprojects.ri.cmu.edu\/2021teame\/wp-content\/uploads\/sites\/50\/2021\/12\/Screen-Shot-2021-12-10-at-5.19.37-PM-2048x661.png 2048w\" sizes=\"auto, (max-width: 767px) 89vw, (max-width: 1000px) 54vw, (max-width: 1071px) 543px, 580px\" \/><\/figure>\n\n\n\n<p><strong><mark style=\"background-color:#ffffff\" class=\"has-inline-color has-vivid-cyan-blue-color\">Presentation Links<\/mark><\/strong><\/p>\n\n\n\n<ul class=\"wp-block-list\"><li>NeuRIPS &#8217;21 AI Driving Olympics Workshop [<a href=\"https:\/\/docs.google.com\/presentation\/d\/12rt87NIUE-y5YY6gcFrtonXU1I9iu3JHFht3Nw9s_yE\/edit?usp=sharing\">Link<\/a>]<\/li><\/ul>\n\n\n\n<p><strong><mark style=\"background-color:#ffffff\" class=\"has-inline-color has-vivid-cyan-blue-color\">Panoptic Segmentation<\/mark><\/strong><\/p>\n\n\n\n<p>To improve the panoptic segmentation quality, we made 4 modifications based on the existing Panoptic Segmentation method <a href=\"https:\/\/arxiv.org\/abs\/2011.11964\">DS-Net<\/a>. <\/p>\n\n\n\n<p>The first 3 modifications are illustrated in following two figures:<\/p>\n\n\n\n<figure class=\"wp-block-image\"><img loading=\"lazy\" decoding=\"async\" width=\"1024\" height=\"549\" src=\"https:\/\/mscvprojects.ri.cmu.edu\/2021teame\/wp-content\/uploads\/sites\/50\/2021\/12\/Screen-Shot-2021-12-10-at-5.36.52-PM-1024x549.png\" alt=\"\" class=\"wp-image-119\" srcset=\"https:\/\/mscvprojects.ri.cmu.edu\/2021teame\/wp-content\/uploads\/sites\/50\/2021\/12\/Screen-Shot-2021-12-10-at-5.36.52-PM-1024x549.png 1024w, https:\/\/mscvprojects.ri.cmu.edu\/2021teame\/wp-content\/uploads\/sites\/50\/2021\/12\/Screen-Shot-2021-12-10-at-5.36.52-PM-300x161.png 300w, https:\/\/mscvprojects.ri.cmu.edu\/2021teame\/wp-content\/uploads\/sites\/50\/2021\/12\/Screen-Shot-2021-12-10-at-5.36.52-PM-768x412.png 768w, https:\/\/mscvprojects.ri.cmu.edu\/2021teame\/wp-content\/uploads\/sites\/50\/2021\/12\/Screen-Shot-2021-12-10-at-5.36.52-PM-1536x824.png 1536w, https:\/\/mscvprojects.ri.cmu.edu\/2021teame\/wp-content\/uploads\/sites\/50\/2021\/12\/Screen-Shot-2021-12-10-at-5.36.52-PM.png 1596w\" sizes=\"auto, (max-width: 767px) 89vw, (max-width: 1000px) 54vw, (max-width: 1071px) 543px, 580px\" \/><\/figure>\n\n\n\n<figure class=\"wp-block-image\"><img loading=\"lazy\" decoding=\"async\" width=\"1024\" height=\"565\" src=\"https:\/\/mscvprojects.ri.cmu.edu\/2021teame\/wp-content\/uploads\/sites\/50\/2021\/12\/Screen-Shot-2021-12-10-at-5.37.07-PM-1024x565.png\" alt=\"\" class=\"wp-image-120\" srcset=\"https:\/\/mscvprojects.ri.cmu.edu\/2021teame\/wp-content\/uploads\/sites\/50\/2021\/12\/Screen-Shot-2021-12-10-at-5.37.07-PM-1024x565.png 1024w, https:\/\/mscvprojects.ri.cmu.edu\/2021teame\/wp-content\/uploads\/sites\/50\/2021\/12\/Screen-Shot-2021-12-10-at-5.37.07-PM-300x166.png 300w, https:\/\/mscvprojects.ri.cmu.edu\/2021teame\/wp-content\/uploads\/sites\/50\/2021\/12\/Screen-Shot-2021-12-10-at-5.37.07-PM-768x424.png 768w, https:\/\/mscvprojects.ri.cmu.edu\/2021teame\/wp-content\/uploads\/sites\/50\/2021\/12\/Screen-Shot-2021-12-10-at-5.37.07-PM-1536x847.png 1536w, https:\/\/mscvprojects.ri.cmu.edu\/2021teame\/wp-content\/uploads\/sites\/50\/2021\/12\/Screen-Shot-2021-12-10-at-5.37.07-PM.png 1548w\" sizes=\"auto, (max-width: 767px) 89vw, (max-width: 1000px) 54vw, (max-width: 1071px) 543px, 580px\" \/><\/figure>\n\n\n\n<p>With these 3 modifications, we successfully <strong>achieved SOTA performance <\/strong>in nuScenes. We also <strong>achieved 2<sup>nd <\/sup>performance <\/strong>in SemanticKITTI without integrating Accumulation. <\/p>\n\n\n\n<p>We carefully ablate the performance improvement due to the modifications in the following table.<\/p>\n\n\n\n<figure class=\"wp-block-image size-full\"><img loading=\"lazy\" decoding=\"async\" width=\"918\" height=\"133\" src=\"https:\/\/mscvprojects.ri.cmu.edu\/2021teame\/wp-content\/uploads\/sites\/50\/2021\/12\/image.png\" alt=\"\" class=\"wp-image-132\" srcset=\"https:\/\/mscvprojects.ri.cmu.edu\/2021teame\/wp-content\/uploads\/sites\/50\/2021\/12\/image.png 918w, https:\/\/mscvprojects.ri.cmu.edu\/2021teame\/wp-content\/uploads\/sites\/50\/2021\/12\/image-300x43.png 300w, https:\/\/mscvprojects.ri.cmu.edu\/2021teame\/wp-content\/uploads\/sites\/50\/2021\/12\/image-768x111.png 768w\" sizes=\"auto, (max-width: 767px) 89vw, (max-width: 1000px) 54vw, (max-width: 1071px) 543px, 580px\" \/><figcaption>Ablation Analysis: Validating the improvements on NuScenes Validation Set<\/figcaption><\/figure>\n\n\n\n<p><strong><span class=\"has-inline-color has-vivid-cyan-blue-color\">Issues with the current model<\/span><\/strong><\/p>\n\n\n\n<p>During the experiments, we noticed that one of the most critical issues of the current method is <strong>low recall<\/strong>, compared with SOTA 3D object detector. <\/p>\n\n\n\n<figure class=\"wp-block-image size-full\"><img loading=\"lazy\" decoding=\"async\" width=\"536\" height=\"368\" src=\"https:\/\/mscvprojects.ri.cmu.edu\/2021teame\/wp-content\/uploads\/sites\/50\/2021\/12\/image-1.png\" alt=\"\" class=\"wp-image-133\" srcset=\"https:\/\/mscvprojects.ri.cmu.edu\/2021teame\/wp-content\/uploads\/sites\/50\/2021\/12\/image-1.png 536w, https:\/\/mscvprojects.ri.cmu.edu\/2021teame\/wp-content\/uploads\/sites\/50\/2021\/12\/image-1-300x206.png 300w\" sizes=\"auto, (max-width: 536px) 100vw, 536px\" \/><\/figure>\n\n\n\n<p>The major reasons are that 1) detector tends to give out <strong>redundant<\/strong> boxes and <strong>reject later<\/strong> 2) Segmentor tends to give <strong>one best<\/strong> label for each point.<\/p>\n\n\n\n<p>Therefore, to improve the <strong>recall<\/strong> of our panoptic segmentor, we allowed <strong>more than one predicted labels<\/strong> for each point, instead of doing clustering over the <strong>top one <\/strong>semantic label. This significantly improved the recall of our model. However, the next step is how to improve <strong>precision<\/strong>.<\/p>\n\n\n\n<p>We are still working on improving the <strong>precision<\/strong>, but we did some initial experiments to verify some possible directions. We <strong>trained a &#8220;scorer&#8221; using MLP regressor, to predict confidence score for each predicted 3D bounding box<\/strong>. By plotting the distribution of the predicted scores from this &#8220;scorer&#8221; for positive\/negative samples, we notice that it is able to perform reasonably well because on average it can assign score &gt; 0.5 for positive samples and score &lt; 0.5 for negative samples.<\/p>\n\n\n\n<figure class=\"wp-block-image\"><img loading=\"lazy\" decoding=\"async\" width=\"1024\" height=\"413\" src=\"https:\/\/mscvprojects.ri.cmu.edu\/2021teame\/wp-content\/uploads\/sites\/50\/2021\/12\/Screen-Shot-2021-12-10-at-5.52.37-PM-1024x413.png\" alt=\"\" class=\"wp-image-123\" srcset=\"https:\/\/mscvprojects.ri.cmu.edu\/2021teame\/wp-content\/uploads\/sites\/50\/2021\/12\/Screen-Shot-2021-12-10-at-5.52.37-PM-1024x413.png 1024w, https:\/\/mscvprojects.ri.cmu.edu\/2021teame\/wp-content\/uploads\/sites\/50\/2021\/12\/Screen-Shot-2021-12-10-at-5.52.37-PM-300x121.png 300w, https:\/\/mscvprojects.ri.cmu.edu\/2021teame\/wp-content\/uploads\/sites\/50\/2021\/12\/Screen-Shot-2021-12-10-at-5.52.37-PM-768x310.png 768w, https:\/\/mscvprojects.ri.cmu.edu\/2021teame\/wp-content\/uploads\/sites\/50\/2021\/12\/Screen-Shot-2021-12-10-at-5.52.37-PM.png 1046w\" sizes=\"auto, (max-width: 767px) 89vw, (max-width: 1000px) 54vw, (max-width: 1071px) 543px, 580px\" \/><\/figure>\n\n\n\n<p><strong><span class=\"has-inline-color has-vivid-cyan-blue-color\">From Panoptic Segmentor to Detector<\/span><\/strong><\/p>\n\n\n\n<p>We aim to explore the possibility of extending a Panoptic Segmentor into a 3D Object Detector for a normal 3D LiDAR Object Objection task. To achieve this goal, we designed and trained an MLP regressor capable of outputting <strong>amodal bounding box<\/strong>, based on the instance labels from our Panoptic Segmentor. Here is a definition of amodal bounding box:<\/p>\n\n\n\n<figure class=\"wp-block-image\"><img loading=\"lazy\" decoding=\"async\" width=\"1024\" height=\"346\" src=\"https:\/\/mscvprojects.ri.cmu.edu\/2021teame\/wp-content\/uploads\/sites\/50\/2021\/12\/Screen-Shot-2021-12-10-at-5.04.56-PM-1024x346.png\" alt=\"\" class=\"wp-image-102\" srcset=\"https:\/\/mscvprojects.ri.cmu.edu\/2021teame\/wp-content\/uploads\/sites\/50\/2021\/12\/Screen-Shot-2021-12-10-at-5.04.56-PM-1024x346.png 1024w, https:\/\/mscvprojects.ri.cmu.edu\/2021teame\/wp-content\/uploads\/sites\/50\/2021\/12\/Screen-Shot-2021-12-10-at-5.04.56-PM-300x101.png 300w, https:\/\/mscvprojects.ri.cmu.edu\/2021teame\/wp-content\/uploads\/sites\/50\/2021\/12\/Screen-Shot-2021-12-10-at-5.04.56-PM-768x259.png 768w, https:\/\/mscvprojects.ri.cmu.edu\/2021teame\/wp-content\/uploads\/sites\/50\/2021\/12\/Screen-Shot-2021-12-10-at-5.04.56-PM-1536x519.png 1536w, https:\/\/mscvprojects.ri.cmu.edu\/2021teame\/wp-content\/uploads\/sites\/50\/2021\/12\/Screen-Shot-2021-12-10-at-5.04.56-PM.png 1546w\" sizes=\"auto, (max-width: 767px) 89vw, (max-width: 1000px) 54vw, (max-width: 1071px) 543px, 580px\" \/><\/figure>\n\n\n\n<p>Our first trial is heuristic method which simply calculates the <strong>mean<\/strong> of all points belonging to an <strong>instance <\/strong>as box center. However, it can only output a modal box instead of amodal box. Therefore, we propose a MLP regressor:<\/p>\n\n\n\n<figure class=\"wp-block-image\"><img loading=\"lazy\" decoding=\"async\" width=\"1024\" height=\"478\" src=\"https:\/\/mscvprojects.ri.cmu.edu\/2021teame\/wp-content\/uploads\/sites\/50\/2021\/12\/Screen-Shot-2021-12-10-at-5.11.05-PM-1024x478.png\" alt=\"\" class=\"wp-image-107\" srcset=\"https:\/\/mscvprojects.ri.cmu.edu\/2021teame\/wp-content\/uploads\/sites\/50\/2021\/12\/Screen-Shot-2021-12-10-at-5.11.05-PM-1024x478.png 1024w, https:\/\/mscvprojects.ri.cmu.edu\/2021teame\/wp-content\/uploads\/sites\/50\/2021\/12\/Screen-Shot-2021-12-10-at-5.11.05-PM-300x140.png 300w, https:\/\/mscvprojects.ri.cmu.edu\/2021teame\/wp-content\/uploads\/sites\/50\/2021\/12\/Screen-Shot-2021-12-10-at-5.11.05-PM-768x359.png 768w, https:\/\/mscvprojects.ri.cmu.edu\/2021teame\/wp-content\/uploads\/sites\/50\/2021\/12\/Screen-Shot-2021-12-10-at-5.11.05-PM-1536x718.png 1536w, https:\/\/mscvprojects.ri.cmu.edu\/2021teame\/wp-content\/uploads\/sites\/50\/2021\/12\/Screen-Shot-2021-12-10-at-5.11.05-PM.png 1768w\" sizes=\"auto, (max-width: 767px) 89vw, (max-width: 1000px) 54vw, (max-width: 1071px) 543px, 580px\" \/><\/figure>\n\n\n\n<p>Our proposed MLP regressor performs <strong>much better<\/strong> compared with the heuristic method, but still suffer from cases where the number of points for an instance is very few.<\/p>\n\n\n\n<figure class=\"wp-block-image size-full\"><img loading=\"lazy\" decoding=\"async\" width=\"559\" height=\"105\" src=\"https:\/\/mscvprojects.ri.cmu.edu\/2021teame\/wp-content\/uploads\/sites\/50\/2021\/12\/image-2.png\" alt=\"\" class=\"wp-image-135\" srcset=\"https:\/\/mscvprojects.ri.cmu.edu\/2021teame\/wp-content\/uploads\/sites\/50\/2021\/12\/image-2.png 559w, https:\/\/mscvprojects.ri.cmu.edu\/2021teame\/wp-content\/uploads\/sites\/50\/2021\/12\/image-2-300x56.png 300w\" sizes=\"auto, (max-width: 559px) 100vw, 559px\" \/><\/figure>\n\n\n\n<p>For more information, please refer to our <a href=\"https:\/\/docs.google.com\/presentation\/d\/12rt87NIUE-y5YY6gcFrtonXU1I9iu3JHFht3Nw9s_yE\/edit?usp=sharing\">presentation<\/a> in <strong><a href=\"https:\/\/driving-olympics.ai\/\">NeurIPS &#8217;21 AI Driving Olympics Workshop<\/a><\/strong>. <\/p>\n","protected":false},"excerpt":{"rendered":"<p>Summary In this semester, our progress can be categorized into two folds: (1) Panoptic Segmentation (2) From Panoptic Segmentor to Detector By the end of this semester, we also submitted our ongoing version of Panoptic Segmentor to NeurIPS &#8217;21 AI Driving Olympics Workshop, and ranked TOP 3 in both LiDAR-track and Open-track (we are Team_AX_Semantic). &hellip; <\/p>\n<p class=\"link-more\"><a href=\"https:\/\/mscvprojects.ri.cmu.edu\/2021teame\/fall-2021\/\" class=\"more-link\">Continue reading<span class=\"screen-reader-text\"> &#8220;Fall 2021&#8221;<\/span><\/a><\/p>\n","protected":false},"author":106,"featured_media":0,"parent":0,"menu_order":0,"comment_status":"closed","ping_status":"closed","template":"","meta":{"footnotes":""},"class_list":["post-78","page","type-page","status-publish","hentry"],"yoast_head":"<!-- This site is optimized with the Yoast SEO plugin v28.1 - https:\/\/yoast.com\/product\/yoast-seo-wordpress\/ -->\n<title>Fall 2021 - Panoptic Lidar Segmentation for Autonomous Driving<\/title>\n<meta name=\"robots\" content=\"index, follow, max-snippet:-1, max-image-preview:large, max-video-preview:-1\" \/>\n<link rel=\"canonical\" href=\"https:\/\/mscvprojects.ri.cmu.edu\/2021teame\/fall-2021\/\" \/>\n<meta property=\"og:locale\" content=\"en_US\" \/>\n<meta property=\"og:type\" content=\"article\" \/>\n<meta property=\"og:title\" content=\"Fall 2021 - Panoptic Lidar Segmentation for Autonomous Driving\" \/>\n<meta property=\"og:description\" content=\"Summary In this semester, our progress can be categorized into two folds: (1) Panoptic Segmentation (2) From Panoptic Segmentor to Detector By the end of this semester, we also submitted our ongoing version of Panoptic Segmentor to NeurIPS &#8217;21 AI Driving Olympics Workshop, and ranked TOP 3 in both LiDAR-track and Open-track (we are Team_AX_Semantic). &hellip; Continue reading &quot;Fall 2021&quot;\" \/>\n<meta property=\"og:url\" content=\"https:\/\/mscvprojects.ri.cmu.edu\/2021teame\/fall-2021\/\" \/>\n<meta property=\"og:site_name\" content=\"Panoptic Lidar Segmentation for Autonomous Driving\" \/>\n<meta property=\"article:modified_time\" content=\"2022-04-29T19:30:28+00:00\" \/>\n<meta property=\"og:image\" content=\"https:\/\/mscvprojects.ri.cmu.edu\/2021teame\/wp-content\/uploads\/sites\/50\/2021\/12\/Screen-Shot-2021-12-10-at-5.19.37-PM-1024x330.png\" \/>\n<meta name=\"twitter:card\" content=\"summary_large_image\" \/>\n<meta name=\"twitter:label1\" content=\"Est. reading time\" \/>\n\t<meta name=\"twitter:data1\" content=\"4 minutes\" \/>\n<script type=\"application\/ld+json\" class=\"yoast-schema-graph\">{\"@context\":\"https:\\\/\\\/schema.org\",\"@graph\":[{\"@type\":\"WebPage\",\"@id\":\"https:\\\/\\\/mscvprojects.ri.cmu.edu\\\/2021teame\\\/fall-2021\\\/\",\"url\":\"https:\\\/\\\/mscvprojects.ri.cmu.edu\\\/2021teame\\\/fall-2021\\\/\",\"name\":\"Fall 2021 - Panoptic Lidar Segmentation for Autonomous Driving\",\"isPartOf\":{\"@id\":\"https:\\\/\\\/mscvprojects.ri.cmu.edu\\\/2021teame\\\/#website\"},\"primaryImageOfPage\":{\"@id\":\"https:\\\/\\\/mscvprojects.ri.cmu.edu\\\/2021teame\\\/fall-2021\\\/#primaryimage\"},\"image\":{\"@id\":\"https:\\\/\\\/mscvprojects.ri.cmu.edu\\\/2021teame\\\/fall-2021\\\/#primaryimage\"},\"thumbnailUrl\":\"https:\\\/\\\/mscvprojects.ri.cmu.edu\\\/2021teame\\\/wp-content\\\/uploads\\\/sites\\\/50\\\/2021\\\/12\\\/Screen-Shot-2021-12-10-at-5.19.37-PM-1024x330.png\",\"datePublished\":\"2021-12-09T14:17:25+00:00\",\"dateModified\":\"2022-04-29T19:30:28+00:00\",\"breadcrumb\":{\"@id\":\"https:\\\/\\\/mscvprojects.ri.cmu.edu\\\/2021teame\\\/fall-2021\\\/#breadcrumb\"},\"inLanguage\":\"en-US\",\"potentialAction\":[{\"@type\":\"ReadAction\",\"target\":[\"https:\\\/\\\/mscvprojects.ri.cmu.edu\\\/2021teame\\\/fall-2021\\\/\"]}]},{\"@type\":\"ImageObject\",\"inLanguage\":\"en-US\",\"@id\":\"https:\\\/\\\/mscvprojects.ri.cmu.edu\\\/2021teame\\\/fall-2021\\\/#primaryimage\",\"url\":\"https:\\\/\\\/mscvprojects.ri.cmu.edu\\\/2021teame\\\/wp-content\\\/uploads\\\/sites\\\/50\\\/2021\\\/12\\\/Screen-Shot-2021-12-10-at-5.19.37-PM.png\",\"contentUrl\":\"https:\\\/\\\/mscvprojects.ri.cmu.edu\\\/2021teame\\\/wp-content\\\/uploads\\\/sites\\\/50\\\/2021\\\/12\\\/Screen-Shot-2021-12-10-at-5.19.37-PM.png\",\"width\":2330,\"height\":752},{\"@type\":\"BreadcrumbList\",\"@id\":\"https:\\\/\\\/mscvprojects.ri.cmu.edu\\\/2021teame\\\/fall-2021\\\/#breadcrumb\",\"itemListElement\":[{\"@type\":\"ListItem\",\"position\":1,\"name\":\"Home\",\"item\":\"https:\\\/\\\/mscvprojects.ri.cmu.edu\\\/2021teame\\\/\"},{\"@type\":\"ListItem\",\"position\":2,\"name\":\"Fall 2021\"}]},{\"@type\":\"WebSite\",\"@id\":\"https:\\\/\\\/mscvprojects.ri.cmu.edu\\\/2021teame\\\/#website\",\"url\":\"https:\\\/\\\/mscvprojects.ri.cmu.edu\\\/2021teame\\\/\",\"name\":\"Panoptic Lidar Segmentation for Autonomous Driving\",\"description\":\"Students: Abhinav Agarwalla, Xuhua Huang | Advisor: Aljosa Osep, James Hays, Deva Ramanan | Sponsor: Argo AI\",\"potentialAction\":[{\"@type\":\"SearchAction\",\"target\":{\"@type\":\"EntryPoint\",\"urlTemplate\":\"https:\\\/\\\/mscvprojects.ri.cmu.edu\\\/2021teame\\\/?s={search_term_string}\"},\"query-input\":{\"@type\":\"PropertyValueSpecification\",\"valueRequired\":true,\"valueName\":\"search_term_string\"}}],\"inLanguage\":\"en-US\"}]}<\/script>\n<!-- \/ Yoast SEO plugin. -->","yoast_head_json":{"title":"Fall 2021 - Panoptic Lidar Segmentation for Autonomous Driving","robots":{"index":"index","follow":"follow","max-snippet":"max-snippet:-1","max-image-preview":"max-image-preview:large","max-video-preview":"max-video-preview:-1"},"canonical":"https:\/\/mscvprojects.ri.cmu.edu\/2021teame\/fall-2021\/","og_locale":"en_US","og_type":"article","og_title":"Fall 2021 - Panoptic Lidar Segmentation for Autonomous Driving","og_description":"Summary In this semester, our progress can be categorized into two folds: (1) Panoptic Segmentation (2) From Panoptic Segmentor to Detector By the end of this semester, we also submitted our ongoing version of Panoptic Segmentor to NeurIPS &#8217;21 AI Driving Olympics Workshop, and ranked TOP 3 in both LiDAR-track and Open-track (we are Team_AX_Semantic). &hellip; Continue reading \"Fall 2021\"","og_url":"https:\/\/mscvprojects.ri.cmu.edu\/2021teame\/fall-2021\/","og_site_name":"Panoptic Lidar Segmentation for Autonomous Driving","article_modified_time":"2022-04-29T19:30:28+00:00","og_image":[{"url":"https:\/\/mscvprojects.ri.cmu.edu\/2021teame\/wp-content\/uploads\/sites\/50\/2021\/12\/Screen-Shot-2021-12-10-at-5.19.37-PM-1024x330.png","type":"","width":"","height":""}],"twitter_card":"summary_large_image","twitter_misc":{"Est. reading time":"4 minutes"},"schema":{"@context":"https:\/\/schema.org","@graph":[{"@type":"WebPage","@id":"https:\/\/mscvprojects.ri.cmu.edu\/2021teame\/fall-2021\/","url":"https:\/\/mscvprojects.ri.cmu.edu\/2021teame\/fall-2021\/","name":"Fall 2021 - Panoptic Lidar Segmentation for Autonomous Driving","isPartOf":{"@id":"https:\/\/mscvprojects.ri.cmu.edu\/2021teame\/#website"},"primaryImageOfPage":{"@id":"https:\/\/mscvprojects.ri.cmu.edu\/2021teame\/fall-2021\/#primaryimage"},"image":{"@id":"https:\/\/mscvprojects.ri.cmu.edu\/2021teame\/fall-2021\/#primaryimage"},"thumbnailUrl":"https:\/\/mscvprojects.ri.cmu.edu\/2021teame\/wp-content\/uploads\/sites\/50\/2021\/12\/Screen-Shot-2021-12-10-at-5.19.37-PM-1024x330.png","datePublished":"2021-12-09T14:17:25+00:00","dateModified":"2022-04-29T19:30:28+00:00","breadcrumb":{"@id":"https:\/\/mscvprojects.ri.cmu.edu\/2021teame\/fall-2021\/#breadcrumb"},"inLanguage":"en-US","potentialAction":[{"@type":"ReadAction","target":["https:\/\/mscvprojects.ri.cmu.edu\/2021teame\/fall-2021\/"]}]},{"@type":"ImageObject","inLanguage":"en-US","@id":"https:\/\/mscvprojects.ri.cmu.edu\/2021teame\/fall-2021\/#primaryimage","url":"https:\/\/mscvprojects.ri.cmu.edu\/2021teame\/wp-content\/uploads\/sites\/50\/2021\/12\/Screen-Shot-2021-12-10-at-5.19.37-PM.png","contentUrl":"https:\/\/mscvprojects.ri.cmu.edu\/2021teame\/wp-content\/uploads\/sites\/50\/2021\/12\/Screen-Shot-2021-12-10-at-5.19.37-PM.png","width":2330,"height":752},{"@type":"BreadcrumbList","@id":"https:\/\/mscvprojects.ri.cmu.edu\/2021teame\/fall-2021\/#breadcrumb","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https:\/\/mscvprojects.ri.cmu.edu\/2021teame\/"},{"@type":"ListItem","position":2,"name":"Fall 2021"}]},{"@type":"WebSite","@id":"https:\/\/mscvprojects.ri.cmu.edu\/2021teame\/#website","url":"https:\/\/mscvprojects.ri.cmu.edu\/2021teame\/","name":"Panoptic Lidar Segmentation for Autonomous Driving","description":"Students: Abhinav Agarwalla, Xuhua Huang | Advisor: Aljosa Osep, James Hays, Deva Ramanan | Sponsor: Argo AI","potentialAction":[{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https:\/\/mscvprojects.ri.cmu.edu\/2021teame\/?s={search_term_string}"},"query-input":{"@type":"PropertyValueSpecification","valueRequired":true,"valueName":"search_term_string"}}],"inLanguage":"en-US"}]}},"_links":{"self":[{"href":"https:\/\/mscvprojects.ri.cmu.edu\/2021teame\/wp-json\/wp\/v2\/pages\/78","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/mscvprojects.ri.cmu.edu\/2021teame\/wp-json\/wp\/v2\/pages"}],"about":[{"href":"https:\/\/mscvprojects.ri.cmu.edu\/2021teame\/wp-json\/wp\/v2\/types\/page"}],"author":[{"embeddable":true,"href":"https:\/\/mscvprojects.ri.cmu.edu\/2021teame\/wp-json\/wp\/v2\/users\/106"}],"replies":[{"embeddable":true,"href":"https:\/\/mscvprojects.ri.cmu.edu\/2021teame\/wp-json\/wp\/v2\/comments?post=78"}],"version-history":[{"count":26,"href":"https:\/\/mscvprojects.ri.cmu.edu\/2021teame\/wp-json\/wp\/v2\/pages\/78\/revisions"}],"predecessor-version":[{"id":216,"href":"https:\/\/mscvprojects.ri.cmu.edu\/2021teame\/wp-json\/wp\/v2\/pages\/78\/revisions\/216"}],"wp:attachment":[{"href":"https:\/\/mscvprojects.ri.cmu.edu\/2021teame\/wp-json\/wp\/v2\/media?parent=78"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}