360度画像R&Dとサービス開発 / 360-degree image research, development, and service development

Transcript

1. ౓ը૾3%ͱ αʔϏε։ൃ 3*$0)
   $0.1"/:
   -5% )JSPDIJLB 'VKJLJ
2. ౻໦ ߂਌ • ;͖͡ ͻΖ͔ͪ • 4NBSU
   7JTJPOࣄۀຊ෦ॴଐ • ౓ը૾Λ࢖ͬͨαʔϏεʹؔ͢Δ ৽ػೳ։ൃΛ୲౰ • αʔϏεʹू·ΔେྔͷσʔλΛ׆͔͠ɺ ػցֶश΍ը૾ॲཧٕज़Λ࠷େݶ׆༻ • $$4&
   ʹҾ͖ଓ͖ࠓ೥΋ొஃ • ࣗ࡞ΩʔϘʔυɺ%WPSBL഑ྻ 
3.  • ͔͑͞͠ ˜+"9"3*$0) 5)&5"
   PO
   *44
4. ౓Χϝϥͷྺ࢙ Nayar, Shree K. "Catadioptric omnidirectional camera." Proceedings of IEEE computer society conference on computer vision and pattern recognition. IEEE, 1997. http://www.cs.columbia.edu/CAVE/projects/cat_cam_360/ Swaminathan, Rahul, and Shree K. Nayar. Polycameras: Camera clusters for wide angle imaging. No. CUCS-013-99. COLUMBIA UNIV NEW YORK DEPT OF COMPUTER SCIENCE, 2005. 
5. 
6. 3*$0)
   5)&5" • ͢΂ͯͷۭؒͷࣸਅɾಈըΛϫϯλονͰ • ੈքॳͷίϯγϡʔϚʔ޲͚౓Χϝϥ 
7. )JTUPSZ
   PG
   5)&5"      5)&5" 5)&5"
   N 5)&5"
   4 5)&5"
   4$ 5)&5"
   7 ,
   TUJMM ,
   TUJMM ,
   WJEFP !GQT ,
   TUJMM ,
   WJEFP !GQT ,
   TUJMM ,
   WJEFP !GQT ,
   TUJMM ,
   WJEFP !GQT ,
   MJWF
   TUSFBNJOH 4QBDJBM BVEJP "OESPJE
   04   5)&5"
   ; 5)&5"
   4$ ,
   TUJMM ,
   WJEFP !GQT ,
   MJWF
   TUSFBNJOH 4QBDJBM BVEJP "OESPJE
   04
   ,
   TUJMM ,
   WJEFP !GQT 
8. 'BDFCPPL
   :PV5VCF
   -*/&
   ͱ͍ͬͨ ϝσΟΞΛڞ༗Ͱ͖ΔαʔϏεͰͷ ౓ը૾ͷαϙʔτ ෆಈ࢈෺݅ͷ಺ݟ ݐઃݱ৔Ͱͷݱ৔ঢ়گڞ༗ $PNQVUFS
   7JTJPO
   Ͱͷར༻ ϩϘοτ޻ֶͰͷར༻ ౓ը૾ͷ׆༻  $POTVNFS #VTJOFTT "DBEFNJB
9. $POTVNFS 
10. ෆಈ࢈޿ࠂͰͷར༻ ؒऔΓਤ Ωϟϓγϣϯ Ξϊςʔγϣϯ ৔ॴͱ޲͖ ྡͷ෦԰΁ Ҡಈ • όʔνϟϧπΞʔͷधཁ૿ • $07*%
    ͷӨڹ͔ΒϦϞʔ τͰޮՌతͳ޿ࠂ͕ඞཁʹ • ౓ը૾Λߦ͖དྷͰ͖Δ όʔνϟϧπΞʔΛ࢖͏͜ͱ Ͱɺ·ΔͰ಺ݟ͍ͯ͠Δ͔ͷ Α͏ʹ෺݅Λݟֶ • ౓ը૾Ҏ֎ͷ৘ใΛը૾ ্ʹܝࡌ 
11. ݐઃݱ৔Ͱͷར༻ • ࢪ޻ঢ়گͷ؅ཧͷࡱӨճ਺Λݮগ • ҟͳΔ࣌ؒʹࡱӨ͞Εͨ౓ը૾Λฒ΂ͯදࣔ͠ࢪ޻؅ཧ 
12. ݚڀྖҬͰͷར༻ Zou, Chuhang, et al. "Layoutnet: Reconstructing the 3d room layout from a single rgb image." Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. 2018. Wang, Ning-Hsu, et al. "360SD-Net: 360› Stereo Depth Estimation with Learnable Cost Volume.” arXiv preprint arXiv:1911.04460 (2019). 
13. ݚڀྖҬͰͷར༻ Guerrero-Viu, Julia, et al. "What's in my Room? Object Recognition on Indoor Panoramic Images." arXiv preprint arXiv:1910.06138 (2019). Zhang, Chao, et al. "Orientation- aware semantic segmentation on icosahedron spheres.” Proceedings of the IEEE International Conference on Computer Vision. 2019. Eder, Marc, et al. "Tangent Images for Mitigating Spherical Distortion." arXiv preprint arXiv:1912.09390 (2019). 
14. ݚڀ༻σʔληοτ Armeni, Iro, et al. "Joint 2d-3d-semantic data for indoor scene understanding." arXiv preprint arXiv:1702.01105 (2017). Chang, Angel, et al. "Matterport3d: Learning from rgb-d data in indoor environments." arXiv preprint arXiv:1709.06158 (2017). Xiao, Jianxiong, et al. "Recognizing scene viewpoint using panoramic place representation." 2012 IEEE Conference on Computer Vision and Pattern Recognition. IEEE, 2012. 
15. 0VS
    QSPEVDUT UFDIOPMPHJFT
16. 7JTJPO  ϫϯγʔϯΛ੾ΓऔΔͷͰ͸ͳ͘ɺશମΛΩϟϓνϟ͢Δ
17. IUUQTXXXSJDPIUPVSTDPN 
18. 3*$0)
    1SPKFDUT  Progress management Simultaneously capture in 360° and pin the location from the project plan. Construction Manager IUUQTXXXSJDPIDPKQTFSWJDFSJDPIQSPKFDUT
19. αʔϏεΛհͯ͠ूΊͨσʔλ • ͜Ε·Ͱ਺ઍສຕͷ౓ ը૾͕ࣗࣾΫϥ΢υαʔϏ εʹ౤ߘ • ͜ΕΒͷσʔλΛ࠷େݶ׆ ༻ͨ͠ػցֶशͷݚڀ։ൃ • ݱ࣮ʹ͋Δ໰୊ͷղܾ • ౓ը૾ʹ͞ΒͳΔ෇Ճ Ձ஋Λ෇༩  3*$0)
    5PVST1SPKFDUT
20. 3*$0)
    5PVST
    ͷಛ௃  όʔνϟϧπΞʔ "*௒ղ૾ "*ը૾੾Γग़͠
21. "*௒ղ૾ • ղ૾ײͷ޲্ɺϊΠζআڈɺίϯτϥετௐ੔ͳͲಉ࣌ʹ࣮ࢪ  ౓ը૾ͷҰ෦Λ੾Γൈ͖
22. σʔληοτͱωοτϫʔΫ • ౓ը૾ͱ "14$
    ͷΧϝϥ ͰࡱӨͨ͠ը૾ͷϖΞΛେྔ࡞ ੒ֶ͠श • αʔϏεԽΛલఏͱ͠ɺॲཧ଎ ౓ͳͲΛҙࣝ͠ &43("/
    Λ ϕʔεͱͨܰ͠ྔͳωοτϫʔ ΫΛ։ൃ • ͜ΕΒͷݚڀΛ·ͱΊͨ࿦จ͕ $713
    ಺ͷ 8PSLTIPQ
    ʹ࠾୒  "
    QBSU
    PG
    B
    FRVJSFDUBOHVMBS
    JNBHF "14$
    DBNFSB
    JNBHF Zhang, Yupeng, et al. "Toward Real-World Panoramic Image Enhancement." Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops. 2020.
23. "*ը૾੾Γग़͠ • ౓ը૾͔Β޿ࠂʹ;͞Θ͍͠௨ৗը֯ͷը૾Λࣗಈੜ੒ • ϨΠΞ΢τਪఆɺॏཁ౓ਪఆɺ෺ମݕ஌ͳͲͷٕज़Λར༻ 
24. ෆಈ࢈޿ࠂʹٻΊΒΕΔ΋ͷ • ෦԰͕޿͘ݟ͑Δ • ෺ମΛத৺ʹࡱӨ • ܏͍͍ͯͳ͍ • ͳͲͳͲɾɾɾ 
25. ػցֶशͷར༻ 4BMJFODZ
    EFUFDUJPO 0CKFDU
    EFUFDUJPO  -BZPVU
    FTUJNBUJPO ෳ਺ͷػցֶशΛ׆༻͠ɺ޿ࠂʹٻΊΒΕΔ ඼࣭Λ࣋ͬͨը૾Λ࠶ݱ
26. ౓ը૾ಛ༗ͷ໰୊఺ • ౓ͱ͍͏ࢹ໺͔֯Βநग़Ͱ͖Δ৘ใͷଟ͕͞ັྗ • Ұํɺຊདྷ͸ٿ໘Ͱ͋ΔͨΊ࿪Έ͕໰୊  https://meder411.medium.com/360-computer-vision-problems-solutions-and-a-suggestion-for-the-path-ahead-438831c6277b
27. ౓ը૾ͷͨΊͷݚڀ  [1] Coors, Benjamin, Alexandru Paul Condurache, and Andreas Geiger. "Spherenet: Learning spherical representations for detection and classification in omnidirectional images." Proceedings of the European Conference on Computer Vision (ECCV). 2018. [2] Zhang, Chao, et al. "Orientation-aware semantic segmentation on icosahedron spheres." Proceedings of the IEEE International Conference on Computer Vision. 2019. [3] Eder, Marc, et al. "Tangent Images for Mitigating Spherical Distortion." Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 2020. <>ΑΓҾ༻ <>ΑΓҾ༻ <>ΑΓҾ༻
28. ౓ը૾ͷػցֶशͷࠓޙ  [4] Zhang, Chao, et al. "Orientation-aware semantic segmentation on icosahedron spheres." Proceedings of the IEEE International Conference on Computer Vision. 2019. [5] Su, Yu-Chuan, and Kristen Grauman. "Kernel transformer networks for compact spherical convolution." Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. 2019. ౤Өํ๏͝ͱͷ࿪Έʹ దԠͨ͠Χʔωϧͷݚڀɺ ٴͼͦͷԠ༻ ௨ৗͷωοτϫʔΫͰ ղܾ͍ͯ͠ΔλεΫʹ ࣅͨܗʹམͱ͠ࠐΉ 3FT/FU ͳͲ܇࿅ࡁΈϞ σϧͷύϥϝʔλΛ ౓ը૾༻ʹม׵ ౓ը૾ͷͨΊ ͷωοτϫʔΫ %FFQ
    -FBSOJOHͷ దԠೳྗͷߴ͞Λ ׆༻ ௨ৗΧʔωϧͷ స༻ <>ΑΓҾ༻ <>ΑΓҾ༻ <>ΑΓҾ༻
29. ࠓޙ  େྔͷ౓ը૾Λ࠷େݶ׆༻͠ɺ੡඼։ൃ΍ݚڀɺ ଞ෼໺΁ͷԠ༻ɺίϥϘϨʔγϣϯΛՃ଎͍͖ͤͯ͞·͢ɻ