{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/multi-task-spatiotemporal-neural-networks-for","title":"Multi-Task Spatiotemporal Neural Networks for Structured Surface Reconstruction","arxiv_id":"1801.03986","date":"2018-01-11","proceeding":null,"authors":["Mingze Xu","Chenyou Fan","John D Paden","Geoffrey C. Fox","David J. Crandall"],"abstract":"Deep learning methods have surpassed the performance of traditional\ntechniques on a wide range of problems in computer vision, but nearly all of\nthis work has studied consumer photos, where precisely correct output is often\nnot critical. It is less clear how well these techniques may apply on\nstructured prediction problems where fine-grained output with high precision is\nrequired, such as in scientific imaging domains. Here we consider the problem\nof segmenting echogram radar data collected from the polar ice sheets, which is\nchallenging because segmentation boundaries are often very weak and there is a\nhigh degree of noise. We propose a multi-task spatiotemporal neural network\nthat combines 3D ConvNets and Recurrent Neural Networks (RNNs) to estimate ice\nsurface boundaries from sequences of tomographic radar images. We show that our\nmodel outperforms the state-of-the-art on this problem by (1) avoiding the need\nfor hand-tuned parameters, (2) extracting multiple surfaces (ice-air and\nice-bed) simultaneously, (3) requiring less non-visual metadata, and (4) being\nabout 6 times faster.","url_abs":"http://arxiv.org/abs/1801.03986v2","url_pdf":"http://arxiv.org/pdf/1801.03986v2.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"multi-task-spatiotemporal-neural-networks-for","repo_url":"https://github.com/shyam1692/ice-reconstruction","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null}],"tasks":[{"task_slug":"structured-prediction","task_name":"Structured Prediction"},{"task_slug":"surface-reconstruction","task_name":"Surface Reconstruction"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":null,"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}