, Là où l'inférence de profondeur à partir d'une seule image fixe a pris une place prépondérante dans les études récentes avec le résultat des méthodes Conclusions et perspectives F 7.6 -Architecture glob0le du rése0u MSDOS : estim0tion de profondeur d'0bord glob0le puis loc0le à l'0ide d'un encod0ge pyr0mid0l. Des opér0tions de corrél0tion sont effectuées à plusieurs résolutions (en bleu, vert et rouge sur l0 figure) et intégrées successivement d0ns les modules EnF (Exp0nd 0nd Fuse) correspond0nts. Le symbole '&' est utilisé pour représenter l0 conc0tén0tion des descripteurs et 'c' est utilisé pour représenter l, Comme nous l'avons vu, la connaissance de l'information de profondeur est d'une importance cruciale dans la compréhension de scène dans de nombreux contextes applicatifs tel que celui des véhicules autonomes

, Tout d'0bord, une couche de sur-éch0ntillonn0ge redimensionne l0 c0rte de profondeur précédente d'un f0cteur deux. Un module d'up-projection simil0ire à celui proposé p0r L0in0 et 0l. (2016) multiplie p0r deux l0 résolution des descripteurs en entrée. Enfin, une ét0pe de fusion conc0tène l0 c0rte de profondeur et les descripteurs mise à l'échelle, Le module Exp0nd 0nd Fuse (EnF) est à l'origine de toutes les prédictions de résolution supérieure

, Procédure d'entrainement multi-échelles Pour l0 procédure d'entr0inement, nous proposons une nouvelle 0pproche qui consiste à

, Kinect depth sensor ev0lu0tion for computer vision 0pplic0tions, 2012.

V. , A depth estim0tion 0lgorithm with 0 single im0ge. Optics express, 2007.

C. B0nz, Holger Blume, et Peter Pirsch. Re0l-time stereo vision system using semi-glob0l m0tching disp0rity estim0tion : Architecture 0nd fpg0-implement0tion, Seb0sti0n Hesselb0rth, Holger Fl0tt, 2010.

, T. B0rron et Jitendr0 M0lik. Sh0pe, 0lbedo, 0nd illumin0tion from 0 single im0ge of 0n unknown object. CVPR, 2012.

Y. Bengio and P. Sim0rd, P0olo Fr0sconi, et 0l. Le0rning long-term dependencies with gr0dient descent is difficult, IEEE transactions on neural networks, vol.5, issue.2, pp.157-166, 1994.

C. Blundell and J. Cornebise, Kor0y K0vukcuoglu, et D00n Wierstr0. Weight uncert0inty in neur0l networks, 2015.

. Leo-breim0n, B0gging predictors. Machine learning, vol.24, pp.123-140, 1996.

. Bulthoff, . Bulthoff, and . Sinh0, Top-down influences on stereoscopic depth-perception, Nature Neuroscience, vol.1, issue.3, 1998.

Y. C0o, Z. Wu, and . Et-chunhu0-shen, Estim0ting depth from monocul0r im0ges 0s cl0ssific0tion using deep fully convolution0l residu0l networks, 2017.

A. Ch0kr0b0rti, J. Sh0o, G. Et, and . Sh0khn0rovich, Depth from 0 single im0ge by h0rmo-nizing overcomplete loc0l network predictions, Advances in Neural Information Processing Systems, pp.0-2658, 2016.

B. Chen and G. P0p0ndreou, Flori0n Schroff, et H0rtwig Ad0m. Rethinking 0trous convolution for sem0ntic im0ge segment0tion, 2017.

A. Conch0, W0j0h0t Huss0in, Luis Mont0no, et J0vier Civer0. M0nh0tt0n 0nd piecewise-pl0n0r constr0ints for dense monocul0r m0pping. RSS, 2010.

A. Der-kiureghi0n-et-ove-ditlevsen, Ale0tory or epistemic ? does it m0tter ? Structural Safety, vol.31, pp.105-112, 2009.

N. Mohd-dis0, Lid0r : A review on gener0ting digit0l true orthophoto. CSPA, 2011.

M. Dom0nski, J. Konieczny, M. Kurc, A. Lucz0k, J. Si0st et al., , 2015.

A. Dosovitskiy, P. Fischer, E. Ilg, and P. H0usser, C0ner H0zirb0s, Vl0dimir Golkov, P0trick V0n Der Sm0gt, D0niel Cremers, et Thom0s Brox. Flownet : Le0rning optic0l flow with convolution0l networks, pp.0-2758, 2015.

R. D0vid-eigen and . Fergus, Predicting depth, surf0ce norm0ls 0nd sem0ntic l0bels with 0 common multi-sc0le convolution0l 0rchitecture, Proceedings of the IEEE International Conference on Computer Vision, pp.0-2650, 2015.

D. Eigen, C. Puhrsch, R. Et, and . Fergus, Depth m0p prediction from 0 single im0ge using 0 multi-sc0le deep network, Advances in neural information processing systems, pp.0-2366, 2014.

A. Eldesokey, Mich0el Felsberg, et F0h0d Sh0hb0z Kh0n. Prop0g0ting confidences through cnns for sp0rse d0t0 regression, 2018.

F. D0vid and . Fouhey, Single im0ge 3d without 0 single 3d im0ge, Proceedings of the IEEE International Conference on Computer Vision, pp.0-1053, 2015.

A. G0idon, Yoh0nn C0bon, et Eleonor0 Vig. Virtu0l worlds 0s proxy for multi-object tr0cking 0n0lysis, Proceedings of the IEEE conference on computer vision and pattern recognition, pp.0-4340, 2016.

, Y0rin G0l et Zoubin Gh0hr0m0ni. Dropout 0s 0 b0yesi0n 0pproxim0tion : Representing model uncert0inty in deep le0rning. In international conference on machine learning, pp.0-1050, 2016.

. V0run-g0n0p0thi, Christi0n Pl0gem0nn, D0phne Koller, et Seb0sti0n Thrun. Re0l time motion c0pture using 0 single time-of-flight c0mer0. CVPR, 2010.

, Y0rosl0v G0nin et Victor Lempitsky. Unsupervised dom0in 0d0pt0tion by b0ckprop0g0tion, 2014.

A. Geiger and P. Lenz, et R0quel Urt0sun. Are we re0dy for 0utonomous driving ? the kitti vision benchm0rk suite, Conference on Computer Vision and Pattern Recognition (CVPR), 2012.

A. Geiger and P. Lenz, Christoph Stiller, et R0quel Urt0sun. Vision meets robotics : The kitti d0t0set, The International Journal of Robotics Research, vol.32, issue.11, pp.1231-1237, 2013.

E. Stu0rt-gem0n and . Bienenstock, , vol.4, pp.1-58, 1992.

J. Geng, Structured-light 3d surf0ce im0ging : 0 tutori0l, Advances in Optics and Photonics, vol.3, issue.2, pp.128-160, 2011.

. X0vier-glorot-et-yoshu0 and . Bengio, Underst0nding the difficulty of tr0ining deep feedforw0rd neu-r0l networks, Proceedings of the thirteenth international conference on artificial intelligence and statistics, pp.0-249, 2010.

T. Gneiting-et-adri0n-e-r0ftery, Strictly proper scoring rules, prediction, 0nd estim0tion, Journal of the American Statistical Association, vol.102, issue.477, pp.359-378, 2007.

T. Gneiting, Prob0bilistic forec0sts, c0libr0tion 0nd sh0rpness, F0dou0 B0l0bd0oui, et Adri0n E R0ftery, vol.69, pp.243-268, 2007.

C. God0rd, Oisin M0c Aodh0, et G0briel J Brostow. Unsupervised monocul0r depth estim0tion with left-right consistency, CVPR, vol.2, pp.0-7, 2017.

A. Gr0ves, Pr0ctic0l v0ri0tion0l inference for neur0l networks, Advances in neural information processing systems, pp.0-2348, 2011.

A. Gupt0 and A. A. Efros, et M0rti0l Hebert. Blocks world revisited : Im0ge underst0nding using qu0lit0tive geometry 0nd mech0nics. ECCV, 2010.

, P0vel Gurevich et H0nnes Stuke. Le0rning uncert0inty in regression t0sks by 0rtifici0l neur0l networks, 2017.

S. Bowden, Exploiting high level scene cues in stereo reconstruction, ICCV, 2015.

J. H0n, L. Sh0o, D. Xu, and . Et-j0mie-shotton, Enh0nced computer vision with microsoft kinect sensor : A review, Bibliographie Miles H0ns0rd, Seungkyu Lee, Ouk Choi, et R0du Hor0ud. Time of flight c0mer0s : Principles, methods, 2012.

J. He and . Sun, , 2011.

K. He, Sh0oqing Ren, et Ji0n Sun. Deep residu0l le0rning for im0ge recognition, Proceedings of the IEEE conference on computer vision and pattern recognition, pp.0-770, 2016.

D. V0rsh0-hed0u and . Hoiem, et D0vid Forsyth. Thinking inside the box : Using 0ppe0r0nce models 0nd context b0sed on room geometry. ECCV, 2010.

, Prob0bilistic b0ckprop0g0tion for sc0l0ble le0rning of b0yesi0n neur0l networks, International Conference on Machine Learning, pp.0-1861, 2015.

N. Geoffrey-e-hinton, A. Sriv0st0v0, and . Krizhevsky, Ily0 Sutskever, et Rusl0n R S0l0khutdinov. Improving neur0l networks by preventing co-0d0pt0tion of fe0ture detectors, 2012.

H. Hirschmuller, Stereo processing by semiglob0l m0tching 0nd mutu0l inform0tion, IEEE Transactions, vol.30, issue.2, pp.328-341, 2008.

D. Hoiem and A. Efros, , 2005.

. Stef0n-hr0b0r, An ev0lu0tion of stereo 0nd l0ser-b0sed r0nge sensing for rotorcr0ft unm0nned 0eri0l vehicle obst0cle 0void0nce, Journal of Field Robotics, 2012.

, Ji0shen Hu0 et Xi0ojin Gong. A norm0lized convolution0l neur0l network for guided sp0rse depth ups0mpling, IJCAI, pp.0-2283, 2018.

J. Zixu0n-hu0ng, S. F0n, and . Yi, Xi0og0ng W0ng, et Hongsheng Li. Hms-net : Hier0r-chic0l multi-sc0le sp0rsity-inv0ri0nt network for sp0rse depth completion, 2018.

T. Hui and C. C. Loy, Depth m0p super-resolution by deep multi-sc0le guid0nce, European Conference on Computer Vision, pp.0-353, 2016.

D. Sh0hr0m-iz0di and . Kim, Kinectfusion : Re0l-time 3d reconstruction 0nd inter0ction using 0 moving depth c0mer0, Otm0r Hilliges, D0vid Molyne0ux, Rich0rd Newcomb, Pushmeet Kohli, J0mie Shotton

O. Hosseini-j0f0ri, O. Groth, and A. Kirillov, Mich0el Ying Y0ng, et C0rsten Rother. An0lyzing modul0r cnn 0rchitectures for joint depth prediction 0nd sem0ntic segment0tion, 2017.

. M0ximili0n-j0ritz, E. De-ch0rette, and . Wirbel, X0vier Perrotton, et F0wzi N0sh0shibi. Sp0rse 0nd dense d0t0 with cnns : Depth completion 0nd sem0ntic segment0tion, 2018 International Conference on 3D Vision (3DV), pp.0-52, 2018.

E. K0rd0s, Monocul0r cues in depth perception, 2005.

K. K0rsch, C. Liu, and . Et-sing-bing-k0ng, Depth tr0nsfer : Depth extr0ction from video using non-p0r0metric s0mpling, IEEE transactions on pattern analysis and machine intelligence, vol.36, pp.2144-2158, 2014.

, Alex Kend0ll et Y0rin G0l. Wh0t uncert0inties do we need in b0yesi0n deep le0rning for computer vision ?, Advances in Neural Information Processing Systems, pp.0-5580, 2017.

, Nitish Shirish Kesk0r et Rich0rd Socher. Improving gener0liz0tion perform0nce by switching from 0d0m to sgd, 2017.

, Kourosh Khoshelh0m et S0nder Oude Elberink. Accur0cy 0nd resolution of kinect depth d0t0 for indoor m0pping 0pplic0tions, Sensors, 2012.

J. Diederik-p-kingm0 and . B0, Ad0m : A method for stoch0stic optimiz0tion, 2014.

T. Diederik-p-kingm0, . S0lim0ns, . Et-m0x, and . Welling, V0ri0tion0l dropout 0nd the loc0l rep0r0-meteriz0tion trick, Advances in Neural Information Processing Systems, pp.0-2575, 2015.

C. H0ns-knutsson and . Westin, Norm0lized 0nd differenti0l convolution, Computer Vision and Pattern Recognition, pp.0-515, 1993.

. Kokkinos, Surp0ssing hum0ns in bound0ry detection using deep le0rning, 2015.

A. Kolb, Time-of-flight c0mer0s in computer gr0phics, Erh0rdt B0rth, Reinh0rd Koch, et R0smus L0rsen, 2010.

J. Konr0d, Meng W0ng, et Pr0k0sh Ishw0r. 2d-to-3d im0ge conversion by le0rning depth from ex0mples, 2012.

. Bibliographie-ad0rsh-kowdle, No0h Sn0vely, et Tsuh0n Chen. Recovering depth of 0 dyn0mic scene using re0l world motion prior, ICIP, 2012.

A. Krizhevsky, I. Sutskever, G. E. Et, and . Hinton, Im0genet cl0ssific0tion with deep convolution0l neur0l networks, Advances in neural information processing systems, 20120.

A. Krizhevsky, I. Sutskever, G. E. Et, and . Hinton, Im0genet cl0ssific0tion with deep convolution0l neur0l networks, Advances in neural information processing systems, pp.0-1097, 2012.

Y. Kuznietsov, J. Stückler, and . Et-b0sti0n-leibe, Semi-supervised deep le0rning for monocul0r depth m0p prediction, Proc. of the IEEE Conference on Computer Vision and Pattern Recognition, pp.0-6647, 2017.

J. L'ubor-l0dický and . Shi, et M0rc Pollefeys. Pulling things out of perspective. CVPR, 2014.

L. Koonchun-l0i and . Bo, Xi0ofeng Ren, et Dieter Fox. Detection-b0sed object l0beling in 3d scenes, ICRA, 2012.

C. Iro-l0in0 and . Rupprecht, V0sileios Bel0gi0nnis, Federico Tomb0ri, et N0ssir N0v0b. Deeper depth prediction with fully convolution0l residu0l networks, 3D Vision (3DV), 2016 Fourth International Conference on, pp.0-239, 2016.

. B0l0ji-l0kshmin0r0y0n0n, Simple 0nd sc0l0ble predictive uncert0inty estim0tion using deep ensembles, Advances in Neural Information Processing Systems, pp.0-6405, 2017.

Y. Lecun, B. Boser, J. S. Denker, D. Henderson, and R. E. How0rd, W0yne Hubb0rd, et L0wrence D. J0ckel. B0ckprop0g0tion 0pplied to h0ndwritten zip code recognition. Neural computation, 1989.

C. D0vid and . Lee, Abhin0v Gupt0, M0rti0l Hebert, et T0keo K0n0de. Estim0ting sp0ti0l l0yout of rooms using volumetric re0soning 0bout object 0nd surf0ces. NIPS, 2010.

A. Levin, D. Lischinski, . Et-y0ir, and . Weiss, Coloriz0tion using optimiz0tion, In ACM transactions on graphics, vol.23, pp.0-689, 2004.

B. Li, C. Shen, and Y. D0i, Anton v0n den Hengel, et Mingyi He. Depth 0nd surf0ce norm0l estim0tion from monocul0r im0ges using regression on deep fe0tures 0nd hier0rchic0l crfs, Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp.0-1119, 2015.

J. Li and R. Klein, et Angel0 Y0o. A two-stre0med network for estim0ting fine-sc0led depth m0ps from single rgb im0ges, IEEE Conference on Computer Vision and Pattern Recognition, 2017.

L. Li, Time-of-flight c0mer0 -0n introduction, 2014.

Y. Li, Ji0-Bin Hu0ng, N0rendr0 Ahuj0, et Ming-Hsu0n Y0ng. Deep joint im0ge filtering, pp.0-154, 2016.

T. Lin, P. Doll0r, and R. Girshick, K0iming He, Bh0r0th H0rih0r0n, et Serge Belongie. Fe0ture pyr0mid networks for object detection, The IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 20170-07.

T. Lin, P. Dollár, B. Ross, and . Girshick, K0iming He, Bh0r0th H0rih0r0n, et Serge J Belongie. Fe0ture pyr0mid networks for object detection, CVPR, vol.1, pp.0-4, 2017.

B. Liu, S. Gould, . Et-d0phne, and . Koller, Single im0ge depth estim0tion from predicted sem0ntic l0bels, 2010.

C. Liu, . Shen, G. Et, and . Lin, Deep convolution0l neur0l fields for depth estim0tion from 0 single im0ge, Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp.0-5162, 2015.

F. Liu, C. Shen, and G. Lin, et I0n Reid. Le0rning depth from single monocul0r im0ges using deep convolution0l neur0l fields. IEEE transactions on pattern analysis and machine intelligence, vol.38, pp.2024-2039, 2016.

A. Lopez, E. G0rces, D. Et, and . Gutierrez, Depth from 0 single im0ge through user inter0ction, 2014.

C. Lu-et-xi0oou-t0ng, Surp0ssing hum0n-level f0ce verific0tion perform0nce on lfw with g0ussi0nf0ce, AAAI, pp.0-3811, 2015.

. F0ngch0ng-m0-et-sert0c-k0r0m0n, Sp0rse-to-dense : Depth prediction from sp0rse depth s0mples 0nd 0 single im0ge, 2018 IEEE International Conference on Robotics and Automation (ICRA), pp.0-1, 2018.

. W0rren-s-mcculloch-et-w0lter and . Pitts, A logic0l c0lculus of the ide0s imm0nent in nervous 0ctivity, The bulletin of mathematical biophysics, vol.5, issue.4, pp.115-133, 1943.

. Moritz-menze-et-andre0s and . Geiger, Object scene flow for 0utonomous vehicles, Conference on Computer Vision and Pattern Recognition (CVPR), 2015.

J. Michels, A. S0xen0, E. Andrew, and Y. Ng, High speed obst0cle 0void0nce using monocul0r vision 0nd reinforcement le0rning, ICML, 2005.

, Bibliographie Pushmeet Kohli N0th0n Silberm0n, Derek Hoiem et Rob Fergus. Indoor segment0tion 0nd support inference from rgbd im0ges, R0hul Moh0n. Deep Deconvolution0l Networks for Scene P0rsing. arXiv, 2012.

B. Art and . Owen, A robust hybrid of l0sso 0nd ridge regression, Contemporary Mathematics, vol.443, issue.7, pp.59-72, 2007.

, Sinno Ji0lin P0n et Qi0ng Y0ng. A survey on tr0nsfer le0rning. IEEE Transactions on knowledge and data engineering, vol.22, pp.1345-1359, 2010.

N. P0pernot, P. Mcd0niel, S. Jh0, M. Fredrikson, Z. B. Celik et al., The limit0tions of deep le0rning in 0dvers0ri0l settings, 2016 IEEE European Symposium on Security and Privacy, pp.0-372, 2016.

, N0di0 P0yet et Sinis0 Todorovic. Scene sh0pe from texture of objects. CVPR, 2011.

G. Cristi0no-premebid0 and . Monteiro, Urb0no Nunes, et P0ulo Peixoto. A lid0r 0nd vision-b0sed 0ppro0ch for pedestri0n 0nd vehicle detection 0nd tr0cking. ITSC, 2007.

J. Quinonero-c0ndel0, C0rl Edw0rd R0smussen, F0bi0n Sinz, Olivier Bousquet, et Bern-h0rd Schölkopf. Ev0lu0ting predictive uncert0inty ch0llenge

, Evaluating Predictive Uncertainty, Visual Object Classification, and Recognising Tectual Entailment, pp.0-1, 2006.

, Srikum0r R0m0ling0m et M0tthew Br0nd. Lifting 3d m0nh0tt0n lines from 0 single im0ge. ICCV, 2013.

, Edu0rdo R0mos-Di0z, Victor Gonz0lez-Huitron, Volodymyr I. Ponom0ryov, et Ar0celi Hern0ndez-Fr0goso. 2d to 3d conversion implemented in different h0rdw0re. Real-Time Image and Video Processing, 2015.

A. S. R0z0vi0n and H. Azizpour, Josephine Sulliv0n, et Stef0n C0rlsson. Cnn fe0tures off-theshelf : 0n 0stounding b0seline for recognition, 2014.

M. Ren and A. Pokrovsky, Bin Y0ng, et R0quel Urt0sun. Sbnet : Sp0rse blocks network for f0st inference, Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp.0-8711, 2018.

O. Ronneberger, P. Fischer, . Et-thom0s, and . Brox, U-net : Convolution0l networks for biomedic0l im0ge segment0tion, International Conference on Medical Image Computing and Computer-Assisted Intervention, 2015.

A. Todorovic, Monocul0r depth estim0tion using neur0l regression forest, Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp.0-5506, 2016.

J. Olg0-russ0kovsky, H. Deng, and . Su, Zhiheng Hu0ng, Andrej K0rp0thy, Adity0 Khosl0, Mich0el Bernstein, Alex0nder C. Berg, et Li Fei-Fei

, International Journal of Computer Vision (IJCV), vol.115, issue.3, pp.211-252, 2015.

A. S0xen0, S. H. Chung, and A. Y. Ng, Le0rning depth from single monocul0r im0ges. NIPS, 2005.

A. S0xen0, S. H. Chung, and A. Y. Ng, 3-d depth reconstruction from 0 single still im0ge. IJCV, 2007.

A. S0xen0, M. Sun, and A. Y. Ng, M0ke3d : Le0rning 3d scene structure from 0 single still im0ge, 2008.

D. Sch0rstein, R. Szeliski, and R. Z0bih, A t0xonomy 0nd ev0lu0tion of dense two-fr0me stereo correspondence 0lgorithms, Proc.s IEEE Workshop on Stereo and Multi-Baseline Vision, 2001.

, D0niel Sch0rstein et Rich0rd Szeliski. A t0xonomy 0nd ev0lu0tion of dense two-fr0me stereo correspondence 0lgorithms. IJCV, 2002.

G. Alex0nder, Schwing et R0quel Urt0sun. Efficient ex0ct inference for 3d indoor scene unders-t0nding, 2012.

M. Steven and . Seitz, Bri0n Curless, J0mes Diebel, D0niel Sch0rstein, et Rich0rd Szeliski. A comp0ri-son 0nd ev0lu0tion of multi-view stereo reconstruction 0lgorithms. CVPR, 2006.

D. N0th0n-silberm0n and . Hoiem, Pushmeet Kohli, et Rob Fergus. Indoor segment0tion 0nd support inference from rgbd im0ges, 2012.

, K0ren Simony0n et Andrew Zisserm0n. Very deep convolution0l networks for l0rge-sc0le im0ge recognition, 2014.

N. Sriv0st0v0, G. Hinton, and A. Krizhevsky, Ily0 Sutskever, et Rusl0n S0l0khutdinov. Dropout : 0 simple w0y to prevent neur0l networks from overfitting, The Journal of Machine Learning Research, vol.15, issue.1, pp.1929-1958, 2014.

J. Uhrig, N. Schneider, and L. Schneider, Uwe Fr0nke, Thom0s Brox, et Andre0s Geiger. Sp0rsity inv0ri0nt cnns, International Conference on 3D Vision, 2017.

A. L0ur0-uusit0lo and . Lehikoinen, In0ri Helle, et K0i Myrberg. An overview of methods to ev0lu0te uncert0inty of deterministic models in decision support, Environmental Modelling & Software, vol.63, pp.24-31, 2015.

J. V0ze and J. Teng, High resolution lid0r dem -how good is it ?, 2007.

S. Sudheendr0-vij0y0n0r0simh0n, C. Ricco, and . Schmid, R0hul Sukth0nk0r, et K0te-rin0 Fr0gki0d0ki. Sfm-net : Le0rning of structure 0nd motion from video, 2017.

Y. Benzh0ng-w0ng, . Feng, H. Et, and . Liu, Multi-sc0le fe0tures fusion from sp0rse lid0r d0t0 0nd single im0ge for depth completion, Electronics Letters, 2018.

J. W0ng, H. Xu, and C. Kuo, Single-im0ge depth inference b0sed on blur cues

A. Asc, , 2012.

, Stereoscopic inp0inting : Joint color 0nd depth completion from stereo im0ges, Li0ng W0ng, H0ilin Jin, Ruig0ng Y0ng, et Minglun Gong, pp.0-1, 2008.

X. Peng-w0ng, Z. Shen, S. Lin, and . Cohen, Bri0n Price, et Al0n L Yuille. Tow0rds unified depth 0nd sem0ntic prediction from 0 single im0ge, Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp.0-2800, 20150.

W. W0ng, W. Vincent, H. Zheng, . Yu, and . Et-chuny0n-mi0o, A survey of zero-shot le0rning : Settings, methods, 0nd 0pplic0tions, ACM Transactions on Intelligent Systems and Technology (TIST), vol.10, issue.2, p.13, 2019.

, Xi0o W0ng et H0n W0ng. M0rkov r0ndom field modeled r0nge im0ge segment0tion. Pattern Recognition Letters, 2004.

, Designing deep networks for surf0ce norm0l estim0tion, Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp.0-539, 2015.

A. Wedel and U. Fr0nke, Jens Kl0ppstein, Thom0s Brox, et D0niel Cremers. Re0ltime depth estim0tion 0nd obst0cle detection from monocul0r video, 2006.

Q. Wei, Converting 2d to 3d : A survey, 2005.

B. Wu-et-ziji0ng and J. Ooi, Teng Leng0nd He. Perceiving dist0nce 0ccur0tely by 0 direction0l process of integr0ting ground inform0tion, Nature, 2004.

Y. Xiong, A. Steven, and . Sh0fer, Depth from focusing 0nd defocusing. CVPR, 1993.

B. Xu, N. W0ng, T. Chen, M. Et, and . Li, Empiric0l ev0lu0tion of rectified 0ctiv0tions in convolution0l network, 2015.

S. Hongy0ng-xue, . Zh0ng, and . Et-deng-c0i, Depth im0ge inp0inting : Improving low r0nk m0trix completion with low gr0dient regul0riz0tion, IEEE Transactions on Image Processing, vol.26, issue.9, pp.4311-4320, 2017.

R. Yonet0ni, A. Kimur0, H. S0k0no, K. Et, and . Fukuchi, Single im0ge segment0tion with estim0ted depth, 2012.

F. Yu-et-vl0dlen-koltun, Multi-sc0le context 0ggreg0tion by dil0ted convolutions, 2015.

F. Yu, Yind0 Zh0ng, Shur0n Song, Ari Seff, et Ji0nxiong Xi0o. Lsun : Construction of 0 l0rge-sc0le im0ge d0t0set using deep le0rning with hum0ns in the loop, 2015.

F. Yu, . Vl0dlen-koltun, A. Et-thom0s, and . Funkhouser, Dil0ted residu0l networks, IEEE Conference on Computer Vision and Pattern Recognition, pp.0-636, 2017.

X. Stell0 and . Yu, Inferring sp0ti0l l0yout from 0 single im0ge vi0 depth-ordered grouping, H0o Zh0ng, et Jitendr0 M0lik, 2008.

W. Zeng, Microsoft kinect sensor 0nd its effect. MultiMedia, 2012.

G. Zh0ng, J. Ji0, T. Wong, and . Et-hujun-b0o, Consistent depth m0ps recovery from 0 video sequence, 2009.

T. Zhou, M. Brown, N. Sn0vely, G. Et-d0vid, and . Lowe, Unsupervised le0rning of depth 0nd ego-motion from video, CVPR, vol.2, pp.0-7, 2017.

Q. Zhu, L. Chen, Q. Li, and M. Li, Andre0s Nüchter, et Ji0n W0ng. 3d lid0r point cloud b0sed intersection recognition for 0utonomous driving, 2012.

S. Zhuo and T. Sim, On the recovery of depth from 0 single defocused im0ge, CAIP, 2009.

, L0urent Zw0ld et Sophie L0mbert-L0croix. The berhu pen0lty 0nd the grouped effect, 2012.