computer-vision-survey

Computer Visionの近年の動向のサーベイ

Transcript

1. Computer visionͷۙ೥ͷಈ޲ͷαʔϕΠ ߴ໦ࢤ࿠ 1
2. αʔϕΠͷ໨త 2 Computer vision (CV) ݚڀͷۙ೥ͷಈ޲Λ஌Γ͍ͨʂ • ֶशख๏Λ஌Γ͍ͨ • ωοτϫʔΫͷมભΛ஌Γ͍ͨ ˠ χϡʔϥϧҎ߱ͷ$7ͷมભ΍͜Ε·Ͱͷಈ޲Λ޿͘ઙ͘঺հ
3. ࠓճ͸࿩͞ͳ͍͜ͱ 3 • ը૾/ಈըੜ੒Ұൠ • ఢରతֶश • ൒ڭࢣ͋Γֶश • ࣗݾڭࢣ͋Γֶश • ݹయతͳίϯϐϡʔλʔϏδϣϯ ͳͲͳͲɽɽ
4. ࠓ೔ͷྲྀΕ 4 ̍ɽλεΫඇಛԽϞσϧʢը૾ೝࣝͷϞσϧʣͷಈ޲ ̎ɽ֤λεΫʹಛԽͨ͠Ϟσϧͷಈ޲ ̏ɽ·ͱΊ
5. ͦͷલʹ 5 ɾਆࢿྉ܈ ɾͪ͜ΒͷࢿྉΛେ͍ʹࢀߟʹ͠·ͨ͠ http://xpaperchallenge.org/cv/ https://github.com/hirokatsukataoka16/cvpaper.challenge-summary
6. ̍ɽλεΫඇಛԽϞσϧͷಈ޲ 6
7. ΞʔΩςΫνϟɾֶश๏ʢը૾ೝࣝʣ 7
8. ࣌ܥྻ 8        
9. AlexNet [Krizhevsky+ NeurIPS 2012] 9 • ը૾ೝࣝίϯϖͰ͋ΔILSVRC2012Ͱѹউ • ਂ૚৞ΈࠐΈχϡʔϥϧωοτϫʔΫ(CNN)ͷ࣌୅ͷນ։͚
10. ࣌ܥྻ 10        
11. ResNet [He+ CVPR 2016] 11 • ILSVRC2015༏উϞσϧ • Skip connectionͷಋೖͰ152૚΋ͷ௒ਂ૚CNNͷֶश͕Մೳʹ • Ҏ߱ͷը૾ೝࣝͷϞσϧ͸جຊతʹResNetͷվྑ
12. ࣌ܥྻ 12        
13. ResNext [Xie+ CVPR 2017] 13 • ೖྗΛ෼ذͤͯ͞ෳ਺ͷωοτϫʔΫͰॲཧ͠ɼͦͷ݁ՌΛ଍͠߹ΘͤΔ
14. WideResNet [Zagoruyko+ 2017] 14 • ਂ͞Λઙͯ͘͠෯Λ޿ͨ͘͠ResNet
15. ࣌ܥྻ 15        
16. PyramidNet [Han+ CVPR 2017] 16 • DownsamplingΛ༻͍Δࡍͷٸܹͳ૚෯૿ՃʹΑΔਫ਼౓ྼԽΛ๷͙ͨΊɼ શମͰগͣͭ͠૚ͷ෯Λେ͖͘͢Δ
17. SENet [Hu+ CVPR 2018] 17 • ૚΁ͷೖྗΛѹॖͨ͠΋ͷΛχϡʔϥϧωοτͰม׵͠ɼ͜ΕΛ༻͍ͯ ೖྗΛॏΈ෇͚Δ
18. DenseNet [Huang+ CVPR 2017 (best paper)] 18 • ֤૚͸ͦͷલͷ͢΂ͯͷ૚ͱskip connectionͰͭͳ͕Δ
19. MobileNet v1-3 [Howard+ 2017, Sandler+ 2018, Howard+ 2019] 19 • ۭؒํ޲ͷΈͷ৞ΈࠐΉdepthwise convolutionͱ νϟωϧํ޲ͷΈ৞ΈࠐΉpointwise convolutionͰ৞ΈࠐΈͷܰྔԽ
20. PNASNet [Liu+ 2017] 20 • Neural architecture search (NAS)ͷ݁ՌಘΒΕͨϞσϧ • CNNશମͰ͸ͳ͘ෳ਺ͷCNNϒϩοΫ͔ΒͳΔʮηϧʯΛ୳ࡧ • ୯७ͳ΋ͷ͔Βঃʑʹෳࡶͳ΋ͷ΁ͱ୳ࡧΛߦ͏
21. ࣌ܥྻ 21        
22. EfficientNet [Tan&Le ICML 2019] 22 • ͜Ε·Ͱͷ༷ʑͳϞσϧͷεέʔϧΞοϓख๏ͷશ෦ͷͤ
23. Noisy Student Training [Xie+ CVPR 2020] 23 • ֶशࡁΈੜెΛڭࢣͱͯ͠ɼॱ࣍େ͖ͳੜెΛֶश͢Δࣗݾڭࢣ͋Γֶश • ੜెʹϊΠζΛ෇Ճ͢Δ͜ͱͰਫ਼౓ʹՃ͑ͯؤ݈ੑ΋޲্
24. BiT [Xie+ Kolesnikov 2019] 24 • ໿10ԯύϥϝʔλͷ௒େن໛ϞσϧͰࣄલֶश • సҠઌͷσʔλ͕গͳͯ͘΋͏·͍͘͘
25. ࣌ܥྻ 25        
26. Vision Transformer (ViT) [Dosovitskiy+ ICLR 2021] 26 • TransformerͰը૾ೝࣝͷSOTA
27. ̎ɽ֤λεΫʹಛԽͨ͠Ϟσϧͷಈ޲ 27
28. ෺ମݕग़ 28
29. Ұൠ෺ମݕग़ 29 [https://pjreddie.com/media/files/papers/YOLOv3.pdf] • ը૾தͷ෺ମͷΫϥεͱҐஔΛ౰ͯΔ
30. ࣌ܥྻ 30 [Zou+ 2020 Object Detection in 20 Years: A Survey]
31. R-CNN [Girshick+ CVPR 2014] 31 • ΦϒδΣΫτ͕ଘࡏ͢ΔީิྖҬΛ੾Γग़͠CNNͰಛ௃நग़
32. Fast R-CNN [Girshick ICCV 2015] 32 • ·ͣը૾ͷಛ௃ϚοϓΛ࡞੒͠ɼީิྖҬ (ROI) Λಛ௃Ϛοϓ্ʹࣹӨ • ΦϒδΣΫτͷ෼ྨͱό΢ϯσΟϯάϘοΫεͷճؼ΋NNͰߦ͏ • ֤ީิྖҬ͝ͱͰ͸ͳ֤͘ը૾͝ͱʹ৞ΈࠐΊ͹Α͘ͳΓɼߴ଎Խ
33. Faster R-CNN [Ren+ NeurIPS 2015] 33 • ީิྖҬ (ROI) ͷఏҊ·ͰؚΊͯend-to-endʹֶश
34. YOLO v1-4 [Redmon+ CVPR 2016, CVPR 2017, 2018, Bochkovskiy+ 2020] 34 • ෺ମݕग़ͱ෺ମࣝผΛҰؾ௨؏ʹߦ͏one-stageͷख๏ • Ϋϥε֬཰ɼ֬৴౓ɼό΢ϯσΟϯάϘοΫεͷ৘ใΛग़ྗ
35. SSD [Liu+ ECCV 2016] 35 • YOLOಉ༷one-stageͷख๏ • ༧Ίෳ༻ҙͨ͠਺ͷό΢ϯσΟϯάϘοΫεຖʹਪ࿦ • ֤૚ͷಛ௃Ϛοϓ͔Βಛ௃நग़͢Δ͜ͱͰ༷ʑͳεέʔϧͰ෺ମݕग़
36. RetinaNet [Lin+ ICCV 2017] 36 • ForegroundͱbackgroundͷΫϥεෆۉߧ͕one-stage๏͕ੑೳͰtwo- stage๏ʹྼΔཧ༝Ͱ͋Δ͜ͱΛࢦఠ • ΫϥεෆۉߧʹରԠ͢ΔͷͨΊͷFocal LossͷఏҊʹΑΓɼ1-stageͳ ͕Βߴ͍ਫ਼౓ͷ෺ମೝࣝΛ࣮ݱ • ϕʔεͷΞʔΩςΫνϟʔʹޙड़ͷFeature Pyramid NetworkΛ࢖༻
37. FCOS [Tian+ ICCV 2019] 37 • RetinaNetͷվྑ൛ • ෺ମͷத৺ͷਪఆΛ௥ՃͰߦ͍ɼΞϯΧʔϑϦʔͳ෺ମݕग़Λ࣮ݱ
38. Bridging the Gap Between Anchor-based and Anchor-free Detection [Zhang+ 2019] 38 • Anchor-basedͱancho-freeͷҧ͍͸ɼෛྫͱਖ਼ྫͷબ୒ͷҧ͍
39. ηάϝϯςʔγϣϯ 39
40. ηάϝϯςʔγϣϯ 40 [https://arxiv.org/pdf/1706.05587.pdf] • ֤ϐΫηϧຖʹ෺ମͷΫϥε/എܠͷࣝผΛ͢Δ
41. ࣌ܥྻ 41 [Minaee+ 2020 Image Segmentation Using Deep Learning: A Survey]
42. FCN [Long+ CVPR 2015] 42 • CNNͷग़ྗ૚΋৞ΈࠐΈ૚ʹ͢Δ͜ͱͰɼώʔτϚοϓΛग़ྗ
43. SegNet [Badrinarayanan+ 2015] 43 • શͯ৞ΈࠐΈ૚ͷΤϯίʔμͱσίʔμ͔ΒͳΔωοτϫʔΫ • σίʔμΛ༻͍Δ͜ͱͰDeconvolution΋ஈ֊తʹߦ͑Δ
44. U-Net [Ronneberger+ MICCAI 2015] 44 • Τϯίʔμͷಛ௃දݱΛskip connectionͰσίʔμʹίϐʔͯ͠౉͢
45. DeepLab v1-3 [Chen+ TPAMI 2017] 45 • Down samplingΛͳ͘͠ɼdilated convolutionͱ૒ઢܗิؒΛ૊Έ߹Θ ͤΔ͜ͱͰߴղ૾౓ͳηάϝϯςʔγϣϯΛ࣮ݱ [Cui+ Remote Sens.2019]
46. FastFCN [Wu+ 2019] 46 • Joint Pyramid Upsampling (JPU) ͷಋೖͰdilated convolutionʹൺ΂ͯ ܭࢉίετΛେ෯ʹ࡟ݮ
47. Mask R-CNN [He+ ICCV 2017] 47 • Bounding boxͷ༧ଌʹՃ͑ͯΫϥεͷϚεΫ΋༧ଌ͢ΔFaster R-CNN • RoIPoolʹ୅ΘΔRoIAlignͷಋೖͰྖҬ෼ׂͳͲ΋Մೳʹ
48. PSPNet [Zhao+ CVPR 2017] 48 • ༷ʑͳεέʔϧͷϓʔϦϯάʹΑΓϚϧνεέʔϧͳಛ௃දݱΛ֫ಘ
49. FPN [Lin+ CVPR 2017] 49 • CNNͷ֊૚ੑΛར༻֤͠֊૚Ͱ༧ଌͯ͠Ϛϧνεέʔϧͳಛ௃Λ֫ಘ • ग़ྗ૚ʹ͍ۙಛ௃Λೖྗ૚ʹ͍ۙଆʹ΋఻͑Δ͜ͱͰɼઙ͍૚Ͱ΋༗ ҙຯͳಛ௃நग़͕Մೳ
50. Visual Question Answering 50
51. Visual Question Answering 51 [https://arxiv.org/pdf/1505.00468.pdf] • ը૾ʹର͢Δ࣭໰จ΁ͷԠ౴
52. ࣌ܥྻ 52 [Srivastava+ 2020 Visual Question Answering using Deep Learning: A Survey and Performance Analysis]
53. σʔληοτ 53 [Srivastava+ 2020 Visual Question Answering using Deep Learning: A Survey and Performance Analysis]
54. VQA [Agrawal+ ICCV 2015] 54 • LSTMͰ࣭໰จΛɼCNNͰը૾ΛຒΊࠐΜͰಛ௃දݱΛ࡞੒
55. Stacked Attention Networks [Yang+ CVPR 2016] 55 • CNNಛ௃ྔʹଟஈ֊ͷattentionΛ͔͚ͯஈ֊తʹର৅ΛߜΓࠐΉ
56. Embodied Question Answering [Das+ CVPR 2018] 56 • ࣭໰͕༩͑ΒΕΔͱɼΤʔδΣϯτ͸γϛϡϨʔγϣϯۭؒ಺Ͱߦಈ Λͱͬͯ౴͑Λݟ͚ͭΔ
57. CLEVR [Johnson+ CVPR 2017] 57 • VQAͷͨΊͷσʔληοτ • ࿦ཧతͳਪ࿦͕ඞཁͱ͞ΕΔ
58. ಈըೝࣝ 58
59. ࣌ܥྻ 59 [Zhu+ 2020 A Comprehensive Study of Deep Video Action Recognition]
60. σʔληοτ 60 [Zhu+ 2020 A Comprehensive Study of Deep Video Action Recognition]
61. ෼ྨ 61 [Zhu+ 2020 A Comprehensive Study of Deep Video Action Recognition]
62. 3D CNN (C3D) [Tran+ ICCV 2015] 62 • 3࣍ݩ৞ΈࠐΈΛ༻͍Δ͜ͱͰ࣌ؒํ޲ͷಛ௃΋දݱ
63. (2+1)D CNN [Tran+ CVPR 2018] 63 • Ұͭͷ૚ͰҰؾʹ࣌ؒํ޲·Ͱ৞ΈࠐΉͷͰ͸ͳ͘ɼ·ۭͣؒํ޲ʹ ৞ΈࠐΜͩ͋ͱͰ࣌ؒํ޲ʹ৞ΈࠐΉ
64. I3D [Carreira&Zisserman CVPR 2017] 64 • 3D ConvΛੵΈॏͶͨωοτϫʔΫ
65. Non-local [Wang+ CVPR 2018] 65 • AttentionʹΑΔॏΈ෇͚ͰɼେҬతͳ৘ใΛՃຯ • ͋ΔҐஔͷ஋Λͦͷଞͷ͢΂ͯͷҐஔͷಛ௃ͷॏΈ෇͖࿨Ͱදݱ
66. SlowFast Networks [Feichtenhofer+ ICCV 2019] 66 • ௿ϑϨʔϜϨʔτͰۭؒಛ௃ΛɼߴϑϨʔϜϨʔτͰ࣌ؒಛ௃Λଊ͑Δ
67. ࢟੎ਪఆ 67
68. ෼ྨ 68 [Chen+ 2020 Monocular Human Pose Estimation: A Survey of Deep Learning-based Methods] [Zheng+ 2020 Deep Learning-Based Human Pose Estimation: A Survey]
69. Convolutional Pose Machines [Wei+ CVPR 2016] 69 • ଟஈ֊ͷ༧ଌʹΑΓɼ֤਎ମ෦Ґͷਪఆਫ਼౓ΛߴΊΔ
70. Part Affinity Fields [Cao+ CVPR 2017] 70 • ࢛ࢶͷҐஔͱ޲͖ΛຒΊࠐΉϕΫτϧ৔Λ༻͍ͨ࢟੎ਪఆ
71. HRNet [Sun+ CVPR 2019] 71 • Sub-networkΛ௥Ճ͢Δ͜ͱͰશମͷղ૾౓Λམͱͣ࢟͞੎ਪఆ͕Մೳ
72. 3D 72
73. ෼ྨ 73 [Ahmed+ 2020 A survey on Deep Learning Advances on Different 3D Data Representations]
74. 3D ఺܈ 74
75. ࣌ܥྻ 75 [Guo+ 2020 Deep Learning for 3D Point Clouds: A Survey]
76. ෼ྨ 76 [Guo+ 2020 Deep Learning for 3D Point Clouds: A Survey]
77. σʔληοτ 77 [Guo+ 2020 Deep Learning for 3D Point Clouds: A Survey]
78. PointNet [Qi+ CVPR 2017] 78 • ఺܈σʔλΛೖྗͱ͠ɼճస΍ॱংͷม׵ͳͲͷૢ࡞ʹରͯ͠ෆมͳಛ ௃Λग़ྗ͢ΔωοτϫʔΫ
79. PointNet++ [Qi+ NeurIPS 2017] 79 • PointNet͸ہॴతͳ৘ใΛ͏·͘र͍͑ͯͳ͔͕ͬͨɼPointNetΛ֊૚త ʹద༻͢Δ͜ͱͰ͜ΕʹରԠ
80. Dynamic Graph CNN [ACMTG+ 2019] 80 • ֤఺ͱͦͷۙ๣ͷؔ܎Λදݱͨ͠Τοδಛ௃Λͭ͘Δ৞ΈࠐΈͷఏҊ
81. VoxelNet [Zhou+ CVPR 2018] 81 • ఺܈σʔλΛvoxelʹ੾Γ෼͚ɼ֤ϘΫηϧ୯ҐͰಛ௃දݱͷຒΊࠐΈ • 3D఺܈෺ମೝࣝͷਫ਼౓޲্
82. 3D ϝογϡ 82
83. Heat Diffusion Equation 83 • ۂ໘ʢϦʔϚϯଟ༷ମʣ্Ͱͷ೤֦ࢄΛߟ͑Δ [Bronstein+ 2016 Geometric deep learning: going beyond Euclidean data]
84. Geodesic CNN [Masci+ ICCV 2015] 84 • ඇϢʔΫϦουଟ༷ମʹ΋ରԠՄೳͳCNNͷఏҊ • ֤఺Ͱۃ࠲ඪΛߟ͑Δ
85. Anisotropic CNN [Boscaini+ NeurIPS 2016] 85 • ඇ౳ํͳ೤ΧʔωϧΛߟ͑Δ͜ͱͰہॴతͳදݱΛΑΓΑ͘நग़ [Bronstein+ 2016 Geometric deep learning: going beyond Euclidean data]
86. Monet [Monti+ CVPR 2017] 86 • ͜Ε·ͰͷඇϢʔΫϦουCNNͷҰൠԽ • ࠲ඪͷҰൠԽ • ݻఆͷΧʔωϧͰ͸ͳֶ͘शՄೳͳΧʔωϧΛ࢖͍ɼΧʔωϧͷҰൠԽ
87. 3D ඍ෼ՄೳϨϯμϥʔ 87
88. ඍ෼ՄೳϨϯμϥʔ 88 % % ϨϯμϦϯά
89. Perspective Transformer Nets [Yan+ NeurIPS 2016] 89 • ϘΫηϧͷඍ෼ՄೳϨϯμϥʔ
90. Neural 3D Mesh Renderer [Monti+ CVPR 2017] 90 • ߴਫ਼౓ͳϝογϡͷඍ෼ՄೳϨϯμϥʔ • ϥελϥΠζ෦෼Λඍ෼Մೳʹͨ͜͠ͱͰٯ఻೻Մೳʹ [https://www.slideshare.net/100001653434308/23d-neural-3d-mesh-renderer-cvpr-2018]
91. Transformers/Attention 91
92. ࣌ܥྻ 92 [Han+ 2021 A Survey on Visual Transformer]
93. ෼ྨ 93 [Han+ 2021 A Survey on Visual Transformer] [Khan+ 2021 Transformers in Vision: A Survey]
94. DETR [Carion+ ECCV 2020] 94 • CNNͰը૾ಛ௃Λநग़ͨ͠ͷͪɼtransformerͰ෺ମೝࣝ
95. iGPT [Chen+ ICML 2020] 95 • ը૾ಛ௃ΛGPT-2Ͱڭࢣͳֶ͠श
96. Vision Transformer (ViT) [Dosovitskiy+ ICLR 2021] 96 • ७ਮͳTransformerͰը૾ೝࣝͷSOTA ࠶ܝ
97. IPT [Chen+ 2020] 97 • ෳ਺ͷλεΫΛಉ࣌ʹߦ͏transformer
98. 98 [https://twitter.com/jaguring1/status/1377710003377725441]
99. 99 [https://www.slideshare.net/cvpaperchallenge/transformer-247407256]
100. ɽ·ͱΊ 100
101. ·ͱΊ 101 • Ϟσϧͷൃల͸ResNetΛϕʔεʹɼෳࡶԽɾେن໛Խɾޮ཰Խ • Vision transformer͕ଓʑొ৔ • جຊతͳcomputer visionͷλεΫʹಛԽͨ͠Ϟσϧ͸ϕϯνϚʔΫ͕ ݻ·͍ͬͯΔ༷ࢠ • 2D → 3DͷྲྀΕ • Ϛϧνεέʔϧͳ৘ใͷ૊ΈࠐΈ͕Α͋͘Δҹ৅ • ࡉ͔͍ςΫχοΫ͕ॏཁͳҹ৅ [https://www.slideshare.net/cvpaperchallenge/cvpr-2020-237139930]
102. ࢀߟࢿྉͳͲ 102
103. ࢀߟࢿྉ 103 • [cvpaper.challenge-summary](https://github.com/hirokatsukataoka16/cvpaper.challenge-summary) • [CVPR 2016 ଎ใ](https://www.slideshare.net/HirokatsuKataoka/cvpr-2016) • [CVPR 2017 ଎ใ](https://www.slideshare.net/cvpaperchallenge/cvpr-2017-78294211) • [CVPR 2018 ଎ใ](https://www.slideshare.net/cvpaperchallenge/cvpr-2018-102878612) • [CVPR 2019 ଎ใ](https://www.slideshare.net/cvpaperchallenge/cvpr-2019) • [CVPR 2020 ଎ใ](https://www.slideshare.net/cvpaperchallenge/cvpr-2020-237139930) • [ಈըೝࣝαʔϕΠv1ʢϝλαʔϕΠ ʣ](https://www.slideshare.net/cvpaperchallenge/v1-232973484) • [Vision and LanguageʢϝλαʔϕΠ ʣ](https://www.slideshare.net/cvpaperchallenge/vision-and-language-232926110) • [৞ΈࠐΈχϡʔϥϧωοτϫʔΫͷݚڀಈ޲](https://www.slideshare.net/ren4yu/ss-84282514) • [ConvNetͷྺ࢙ͱResNetѥछɺ΂ετϓϥΫςΟε](https://www.slideshare.net/ren4yu/convnetresnet) • [৞ΈࠐΈχϡʔϥϧωοτϫʔΫͷߴਫ਼౓Խͱߴ଎Խ](https://www.slideshare.net/ren4yu/ss-145689425) • [࿦จ঺հ: Fast R-CNN&Faster R-CNN](https://www.slideshare.net/takashiabe338/fast-rcnnfaster-rcnn) • [ʲ෺ମݕग़ʳSSD(Single Shot MultiBox Detector)ͷղઆ](https://www.acceluniverse.com/blog/developers/2020/02/SSD.html) • [ʲ෺ମݕग़ख๏ͷྺ࢙ : YOLOͷ঺հʳ](https://qiita.com/cv_carnavi/items/68dcda71e90321574a2b) • [ը૾ೝࣝͱਂ૚ֶश](https://www.slideshare.net/ren4yu/ss-234439652) • [semantic segmentation αʔϕΠ](https://www.slideshare.net/yoheiokawa/semantic-segmentation-141471958) • [Semantic segmentation ৼΓฦΓ](https://speakerdeck.com/motokimura/semantic-segmentation-zhen-rifan-ri) • [[DLྠಡձ]SlowFast Networks for Video Recognition](https://www.slideshare.net/DeepLearningJP2016/dlslowfast-networks-for-video-recognition-202057397) • [ࡾ࣍ݩ఺܈ΛऔΓѻ͏χϡʔϥϧωοτϫʔΫͷαʔϕΠ](https://www.slideshare.net/naoyachiba18/ss-120302579) • [ࡾ࣍ݩ఺܈ΛऔΓѻ͏χϡʔϥϧωοτϫʔΫͷαʔϕΠ Ver. 2](https://speakerdeck.com/nnchiba/point-cloud-deep-learning-survey-ver-2) • [఺܈ਂ૚ֶश Meta-study](https://www.slideshare.net/naoyachiba18/metastudy) • [ୈ̍ճ ࠷৽ͷML,CV,NLP ؔ࿈࿦จಡΈձ PointNet](https://www.slideshare.net/FujimotoKeisuke/point-net) • [ [DLྠಡձ]MeshͱDeep Learning Surface Networks & AtlasNet](https://www.slideshare.net/DeepLearningJP2016/dlmeshdeep-learning-surface-networks-atlasnet) • [࿦จ·ͱΊɿConvolutional Pose Machines](https://qiita.com/masataka46/items/88f1a375ce8a485d9454) • [ίϯϐϡʔλϏδϣϯͷ࠷৽࿦จௐࠪ 2D Human Pose Estimation ฤ](https://engineer.dena.com/posts/2019.11/cv-papers-19-2d-human-pose-estimation/) • [[ୈ2ճ3Dษڧձ ݚڀ঺հ] Neural 3D Mesh Renderer (CVPR 2018)](https://www.slideshare.net/100001653434308/23d-neural-3d-mesh-renderer-cvpr-2018) • [DeepLabʹ୅ΘΓݱࡏͷSOTAͰ͋ΔFastFCN(JPU)ͷ࿦จղઆ](https://qiita.com/kamata1729/items/1b495658a63d76904ac3)