View this talk on talks.cam (opens in new tab)<\/span><\/a>Opens in a new tab<\/span><\/p>\n Frontiers in Artificial Intelligence is a series of public lectures at Microsoft Research Cambridge featuring leading researchers in the field, focusing on the cutting edge topics at the intersection of machine learning, statistics, and artificial intelligence. Students, scientists, and engineers in academia and industry are all welcome to join us for these exciting talks and the opportunity to socialize with the Cambridge AI\/ML community.<\/p>\n 3D reconstruction from multiple 2D images is an inherently ill-posed problem. Prior knowledge is required to resolve ambiguities and probabilistic models are desirable to capture the ambiguities in the reconstructed model. In this talk, I will present two recent results tackling these two aspects. First, I will introduce a probabilistic framework for volumetric 3D reconstruction where the reconstruction problem is cast as inference in a Markov random field using ray potentials. Our main contribution is a discrete-continuous inference algorithm which computes marginal distributions of each voxel’s occupancy and appearance. I will show that the proposed algorithm allows for Bayes optimal predictions with respect to a natural reconstruction loss. I will further demonstrate several extensions which integrate non-local CAD priors into the reconstruction process. In the second part of my talk, I will present a novel framework for deep learning with 3D data called OctNet which enables 3D CNNs on high-dimensional inputs. I will demonstrate the utility of the OctNet representation on several 3D tasks including classification, orientation estimation and point cloud labeling. Finally, I will present an extension of OctNet called OctNetFusion which jointly predicts the space partitioning function with the output representation, resulting in an end-to-end trainable model for volumetric depth map fusion.Opens in a new tab<\/span><\/p>\n Francesco Orabona – Coin Betting for Backprop without Learning Rates and More<\/a><\/p>\n Regina Barzilay – How Can NLP Help Cure Cancer? (opens in new tab)<\/span><\/a><\/p>\n Aapo Hyvarinen – Nonlinear ICA using temporal structure: a principled framework for unsupervised deep learning\u00a0<\/a><\/p>\n Max Welling – Generalizing Convolutions for Deep Learning <\/a>Opens in a new tab<\/span><\/p>\n","protected":false},"excerpt":{"rendered":" 21 Station Road Cambridge CB1 2FB View the whole series on talks.cam (opens in new tab) View this talk on talks.cam (opens in new tab)Opens in a new tab Frontiers in Artificial Intelligence is a series of public lectures at Microsoft Research Cambridge featuring leading researchers in the field, focusing on the cutting edge topics […]<\/p>\n","protected":false},"featured_media":0,"template":"","meta":{"msr-url-field":"","msr-podcast-episode":"","msrModifiedDate":"","msrModifiedDateEnabled":false,"ep_exclude_from_search":false,"_classifai_error":"","msr_startdate":"2017-09-26","msr_enddate":"2017-09-26","msr_location":"Microsoft Research Cambridge UK","msr_expirationdate":"","msr_event_recording_link":"","msr_event_link":"","msr_event_link_redirect":false,"msr_event_time":"13:00","msr_hide_region":true,"msr_private_event":false,"msr_hide_image_in_river":0,"footnotes":""},"research-area":[13556],"msr-region":[239178],"msr-event-type":[197944],"msr-video-type":[],"msr-locale":[268875],"msr-program-audience":[],"msr-post-option":[],"msr-impact-theme":[],"class_list":["post-423669","msr-event","type-msr-event","status-publish","hentry","msr-research-area-artificial-intelligence","msr-region-europe","msr-event-type-hosted-by-microsoft","msr-locale-en_us"],"msr_about":"\n\n 21 Station Road View the whole series on talks.cam (opens in new tab)<\/span><\/a><\/p>\n View this talk on talks.cam (opens in new tab)<\/span><\/a>Opens in a new tab<\/span><\/p>\n Frontiers in Artificial Intelligence is a series of public lectures at Microsoft Research Cambridge featuring leading researchers in the field, focusing on the cutting edge topics at the intersection of machine learning, statistics, and artificial intelligence. Students, scientists, and engineers in academia and industry are all welcome to join us for these exciting talks and the opportunity to socialize with the Cambridge AI\/ML community.<\/p>\n 3D reconstruction from multiple 2D images is an inherently ill-posed problem. Prior knowledge is required to resolve ambiguities and probabilistic models are desirable to capture the ambiguities in the reconstructed model. In this talk, I will present two recent results tackling these two aspects. First, I will introduce a probabilistic framework for volumetric 3D reconstruction where the reconstruction problem is cast as inference in a Markov random field using ray potentials. Our main contribution is a discrete-continuous inference algorithm which computes marginal distributions of each voxel’s occupancy and appearance. I will show that the proposed algorithm allows for Bayes optimal predictions with respect to a natural reconstruction loss. I will further demonstrate several extensions which integrate non-local CAD priors into the reconstruction process. In the second part of my talk, I will present a novel framework for deep learning with 3D data called OctNet which enables 3D CNNs on high-dimensional inputs. I will demonstrate the utility of the OctNet representation on several 3D tasks including classification, orientation estimation and point cloud labeling. Finally, I will present an extension of OctNet called OctNetFusion which jointly predicts the space partitioning function with the output representation, resulting in an end-to-end trainable model for volumetric depth map fusion.Opens in a new tab<\/span><\/p>\n Francesco Orabona – Coin Betting for Backprop without Learning Rates and More<\/a><\/p>\n Regina Barzilay – How Can NLP Help Cure Cancer?<\/a><\/p>\n Aapo Hyvarinen – Nonlinear ICA using temporal structure: a principled framework for unsupervised deep learning\u00a0<\/a><\/p>\n Max Welling – Generalizing Convolutions for Deep Learning <\/a>Opens in a new tab<\/span><\/p>\n","tab-content":[{"id":0,"name":"About","content":"Frontiers in Artificial Intelligence is a series of public lectures at Microsoft Research Cambridge featuring leading researchers in the field, focusing on the cutting edge topics at the intersection of machine learning, statistics, and artificial intelligence. Students, scientists, and engineers in academia and industry are all welcome to join us for these exciting talks and the opportunity to socialize with the Cambridge AI\/ML community.\r\n\r\n![\"\"](\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2017\/09\/portrait_small-210x300.jpg\")
<\/p>\nProbabilistic and Deep Models for 3D Reconstruction<\/span><\/h3>\n
Andreas Geiger – Max Planck Institute for Intelligent Systems<\/h4>\n
\nCambridge
\nCB1 2FB<\/p>\n<\/p>\nProbabilistic and Deep Models for 3D Reconstruction<\/span><\/h3>\n
Andreas Geiger – Max Planck Institute for Intelligent Systems<\/h4>\n
\r\nProbabilistic and Deep Models for 3D Reconstruction<\/span><\/h3>\r\n
Andreas Geiger - Max Planck Institute for Intelligent Systems<\/h4>\r\n3D reconstruction from multiple 2D images is an inherently ill-posed problem. Prior knowledge is required to resolve ambiguities and probabilistic models are desirable to capture the ambiguities in the reconstructed model. In this talk, I will present two recent results tackling these two aspects. First, I will introduce a probabilistic framework for volumetric 3D reconstruction where the reconstruction problem is cast as inference in a Markov random field using ray potentials. Our main contribution is a discrete-continuous inference algorithm which computes marginal distributions of each voxel's occupancy and appearance. I will show that the proposed algorithm allows for Bayes optimal predictions with respect to a natural reconstruction loss. I will further demonstrate several extensions which integrate non-local CAD priors into the reconstruction process. In the second part of my talk, I will present a novel framework for deep learning with 3D data called OctNet which enables 3D CNNs on high-dimensional inputs. I will demonstrate the utility of the OctNet representation on several 3D tasks including classification, orientation estimation and point cloud labeling. Finally, I will present an extension of OctNet called OctNetFusion which jointly predicts the space partitioning function with the output representation, resulting in an end-to-end trainable model for volumetric depth map fusion."},{"id":1,"name":"Past Speakers","content":"Francesco Orabona - Coin Betting for Backprop without Learning Rates and More<\/a>\r\n\r\nRegina Barzilay - How Can NLP Help Cure Cancer?<\/a>\r\n\r\nAapo Hyvarinen - Nonlinear ICA using temporal structure: a principled framework for unsupervised deep learning\u00a0<\/a>\r\n\r\nMax Welling - Generalizing Convolutions for Deep Learning <\/a>"}],"msr_startdate":"2017-09-26","msr_enddate":"2017-09-26","msr_event_time":"13:00","msr_location":"Microsoft Research Cambridge UK","msr_event_link":"","msr_event_recording_link":"","msr_startdate_formatted":"September 26, 2017","msr_register_text":"Watch now","msr_cta_link":"","msr_cta_text":"","msr_cta_bi_name":"","featured_image_thumbnail":null,"event_excerpt":"Frontiers in Artificial Intelligence is a series of public lectures at Microsoft Research Cambridge featuring leading researchers in the field, focusing on the cutting edge topics at the intersection of machine learning, statistics, and artificial intelligence. Students, scientists, and engineers in academia and industry are all welcome to join us for these exciting talks and the opportunity to socialize with the Cambridge AI\/ML community. Probabilistic and Deep Models for 3D Reconstruction Andreas Geiger - Max…","msr_research_lab":[199561],"related-researchers":[],"msr_impact_theme":[],"related-academic-programs":[],"related-groups":[],"related-projects":[],"related-opportunities":[],"related-publications":[],"related-videos":[],"related-posts":[],"_links":{"self":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-event\/423669","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-event"}],"about":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/types\/msr-event"}],"version-history":[{"count":2,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-event\/423669\/revisions"}],"predecessor-version":[{"id":1147149,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-event\/423669\/revisions\/1147149"}],"wp:attachment":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/media?parent=423669"}],"wp:term":[{"taxonomy":"msr-research-area","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/research-area?post=423669"},{"taxonomy":"msr-region","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-region?post=423669"},{"taxonomy":"msr-event-type","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-event-type?post=423669"},{"taxonomy":"msr-video-type","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-video-type?post=423669"},{"taxonomy":"msr-locale","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-locale?post=423669"},{"taxonomy":"msr-program-audience","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-program-audience?post=423669"},{"taxonomy":"msr-post-option","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-post-option?post=423669"},{"taxonomy":"msr-impact-theme","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-impact-theme?post=423669"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}