000075699 001__ 75699
000075699 005__ 20190819101353.0
000075699 0247_ $$2doi$$a10.1145/3130800.3130884
000075699 0248_ $$2sideral$$a104307
000075699 037__ $$aART-2017-104307
000075699 041__ $$aeng
000075699 100__ $$0(orcid)0000-0001-9960-8945$$aMarco, J.$$uUniversidad de Zaragoza
000075699 245__ $$aDeepToF: Off-the-shelf real-time correction of multipath interference in time-of-flight imaging
000075699 260__ $$c2017
000075699 5060_ $$aAccess copy available to the general public$$fUnrestricted
000075699 5203_ $$aTime-of-flight (ToF) imaging has become a widespread technique for depth estimation, allowing affordable off-the-shelf cameras to provide depth maps in real time. However, multipath interference (MPI) resulting from indirect illumination significantly degrades the captured depth. Most previous works have tried to solve this problem by means of complex hardware modifications or costly computations. In this work, we avoid these approaches and propose a new technique to correct errors in depth caused by MPI, which requires no camera modifications and takes just 10 milliseconds per frame. Our observations about the nature of MPI suggest that most of its information is available in image space; this allows us to formulate the depth imaging process as a spatially-varying convolution and use a convolutional neural network to correct MPI errors. Since the input and output data present similar structure, we base our network on an autoencoder, which we train in two stages. First, we use the encoder (convolution filters) to learn a suitable basis to represent MPI-corrupted depth images; then, we train the decoder (deconvolution filters) to correct depth from synthetic scenes, generated by using a physically-based, time-resolved renderer. This approach allows us to tackle a key problem in ToF, the lack of ground-truth data, by using a large-scale captured training set with MPI-corrupted depth to train the encoder, and a smaller synthetic training set with ground truth depth to train the decoder stage of the network. We demonstrate and validate our method on both synthetic and real complex scenarios, using an off-the-shelf ToF camera, and with only the captured, incorrect depth as input.
000075699 536__ $$9info:eu-repo/grantAgreement/ES/MINECO/TIN2016-78753-P$$9info:eu-repo/grantAgreement/ES/MINECO/TIN2014-61696-EXP$$9This project has received funding from the European Union’s Horizon 2020 research and innovation program under grant agreement No H2020 682080-CHAMELEON$$9info:eu-repo/grantAgreement/EC/H2020/682080/EU/Intuitive editing of visual appearance from real-world datasets/CHAMELEON
000075699 540__ $$9info:eu-repo/semantics/openAccess$$aAll rights reserved$$uhttp://www.europeana.eu/rights/rr-f/
000075699 590__ $$a4.384$$b2017
000075699 591__ $$aCOMPUTER SCIENCE, SOFTWARE ENGINEERING$$b3 / 104 = 0.029$$c2017$$dQ1$$eT1
000075699 592__ $$a1.344$$b2017
000075699 593__ $$aComputer Graphics and Computer-Aided Design$$c2017$$dQ1
000075699 655_4 $$ainfo:eu-repo/semantics/article$$vinfo:eu-repo/semantics/acceptedVersion
000075699 700__ $$aHernandez, Q.
000075699 700__ $$0(orcid)0000-0002-8160-7159$$aMuñoz, A.$$uUniversidad de Zaragoza
000075699 700__ $$aDong, Y.
000075699 700__ $$0(orcid)0000-0001-9000-0466$$aJarabo, A.
000075699 700__ $$aKim, M.H.
000075699 700__ $$aTong, X.
000075699 700__ $$0(orcid)0000-0002-7503-7022$$aGutierrez, D.$$uUniversidad de Zaragoza
000075699 7102_ $$15007$$2570$$aUniversidad de Zaragoza$$bDpto. Informát.Ingenie.Sistms.$$cÁrea Lenguajes y Sistemas Inf.
000075699 773__ $$g36, 6 (2017), 219 [12 pp]$$pACM trans. graph.$$tACM TRANSACTIONS ON GRAPHICS$$x0730-0301
000075699 8564_ $$s1372965$$uhttps://zaguan.unizar.es/record/75699/files/texto_completo.pdf$$yPostprint
000075699 8564_ $$s93611$$uhttps://zaguan.unizar.es/record/75699/files/texto_completo.jpg?subformat=icon$$xicon$$yPostprint
000075699 909CO $$ooai:zaguan.unizar.es:75699$$particulos$$pdriver
000075699 951__ $$a2019-08-19-09:51:36
000075699 980__ $$aARTICLE