Supplementary Material for Deep View Synthesis from
SparsePhotometric Images
ZEXIANG XU, University of California, San DiegoSAI BI,
University of California, San DiegoKALYAN SUNKAVALLI, Adobe
ResearchSUNIL HADAP∗, Lab126, AmazonHAO SU, University of
California, San DiegoRAVI RAMAMOORTHI, University of California,
San Diego
3 3D 3x3x3 Conv, stride 2 + GN + ReLU
3 3D 3x3x3 Conv, stride 1 + GN + ReLU
3 Nearest neighbor upsample, + 3D 3x3x3 Conv stride 1 + GN +
ReLU
3 2D 3x3 Conv, stride 1+ Sigmoid 3 3D 3x3x3 Conv, stride
1+Tanh
3 3D 3x3x3 Conv, stride 1+ Sigmoid2D feature map noted with
channel number 3D feature volume noted with channel numberSkip
link
Features from skip links
833
3
3 3
3
3
163 16 16 16 8
3 2D 3x3 Conv, stride 2 + BN + ReLU
3 2D 3x3 Conv, stride 1 + BN + ReLU
3 Nearest neighbor upsample + 2D 3x3 Conv stride 1 + BN +
ReLU
3
3
3
3
3
3
3
9m+1
32 32 64 643
643
64 32 32 16 m+1
3
3
3
3
3
3
3
3
3
5m+1
64 128
256
512
3
512
3
256
128 64 32 32 3
3Fig. 1. The network details of our feature extractor T,
correspondence predictor C and shading predictor S
.ACM Reference Format:Zexiang Xu, Sai Bi, Kalyan Sunkavalli,
Sunil Hadap, Hao Su, and Ravi Ra-mamoorthi. 2019. Supplementary
Material for Deep View Synthesis fromSparse Photometric Images. ACM
Trans. Graph. 38, 4, Article 76 (July 2019),1 page.
https://doi.org/10.1145/3306346.3323007
∗This work was done prior to joining Amazon.
Authors’ addresses: Zexiang Xu, University of California, San
Diego, [email protected]; Sai Bi, University of California, San
Diego, [email protected]; KalyanSunkavalli, Adobe Research,
[email protected]; Sunil Hadap, Lab126, Amazon,[email protected];
Hao Su, University of California, San Diego, [email protected];
Ravi Ramamoorthi, University of California, San Diego,
[email protected].
© 2019 Copyright held by the owner/author(s). Publication rights
licensed to ACM.This is the author’s version of the work. It is
posted here for your personal use. Not forredistribution. The
definitive Version of Record was published in ACM Transactions
onGraphics, https://doi.org/10.1145/3306346.3323007.
ACM Trans. Graph., Vol. 38, No. 4, Article 76. Publication date:
July 2019.
https://doi.org/10.1145/3306346.3323007https://doi.org/10.1145/3306346.3323007
