《计算机应用》唯一官方网站 ›› 2023, Vol. 43 ›› Issue (3): 894-902.DOI: 10.11772/j.issn.1001-9081.2022101589
• 多媒体计算与计算机仿真 • 上一篇
张江峰1,2, 闫涛1,2,3,4(), 陈斌4,5, 钱宇华2,3, 宋艳涛1,2,3
收稿日期:
2022-10-25
修回日期:
2023-01-12
接受日期:
2023-01-16
发布日期:
2023-03-15
出版日期:
2023-03-10
通讯作者:
闫涛
作者简介:
张江峰(1998—),男,山西晋城人,硕士研究生,CCF会员,主要研究方向:深度学习、三维重建基金资助:
Jiangfeng ZHANG1,2, Tao YAN1,2,3,4(), Bin CHEN4,5, Yuhua QIAN2,3, Yantao SONG1,2,3
Received:
2022-10-25
Revised:
2023-01-12
Accepted:
2023-01-16
Online:
2023-03-15
Published:
2023-03-10
Contact:
Tao YAN
About author:
ZHANG Jiangfeng, born in 1998, M. S. candidate. His research interests include deep learning, 3D reconstruction.Supported by:
摘要:
针对现有三维形貌重建模型无法有效融合全局时空信息的问题,设计深度聚焦体积(DFV)模块保留聚焦和离焦的过渡信息,并在此基础上提出全局时空特征耦合(GSTFC)模型提取多景深图像序列的局部与全局的时空特征信息。首先,在收缩路径中穿插3D-ConvNeXt模块和3D卷积层,捕捉多尺度局部时空特征,同时,在瓶颈模块中添加3D-SwinTransformer模块捕捉多景深图像序列局部时序特征的全局关联关系;然后,通过自适应参数层将局部时空特征和全局关联关系融合为全局时空特征,并输入扩张路径引导生成聚焦体积;最后,聚焦体积通过DFV提取序列权重信息,并保留聚焦与离焦的过渡信息,得到最终深度图。实验结果表明,GSTFC在FoD500数据集上的均方根误差(RMSE)相较于最先进的全聚焦深度网络(AiFDepthNet)下降了12.5%,并且比传统的鲁棒聚焦体积正则化的聚焦形貌恢复(RFVR-SFF)模型保留了更多的景深过渡关系。
中图分类号:
张江峰, 闫涛, 陈斌, 钱宇华, 宋艳涛. 全局时空特征耦合的多景深三维形貌重建[J]. 计算机应用, 2023, 43(3): 894-902.
Jiangfeng ZHANG, Tao YAN, Bin CHEN, Yuhua QIAN, Yantao SONG. Multi-depth-of-field 3D shape reconstruction with global spatio-temporal feature coupling[J]. Journal of Computer Applications, 2023, 43(3): 894-902.
模型 | 主干 | 模块 | 特征处理 | MSE | 参数量/106 | 浮点运算量/GFLOPs |
---|---|---|---|---|---|---|
U | 3D-U型主干 | — | 卷积层 | 0.004 23 | 29.160 0 | 534.27 |
U+T | 3D-U型主干 | 3D-SwinTransformer | 卷积层 | 0.003 73 | 35.657 0 | 549.18 |
U+T+DFV | 3D-U型主干 | 3D-SwinTransformer | DFV | 35.660 0 | 549.18 | |
X | 3D-ConvNeXt | — | 卷积层 | 0.005 10 | 9.760 0 | 233.17 |
X+T | 3D-ConvNeXt | 3D-SwinTransformer | 卷积层 | 0.002 21 | 16.251 1 | 248.09 |
X+T+DFV | 3D-ConvNeXt | 3D-SwinTransformer | DFV | 0.000 98 |
表1 消融实验的对比分析
Tab. 1 Comparison analysis of ablation experiments
模型 | 主干 | 模块 | 特征处理 | MSE | 参数量/106 | 浮点运算量/GFLOPs |
---|---|---|---|---|---|---|
U | 3D-U型主干 | — | 卷积层 | 0.004 23 | 29.160 0 | 534.27 |
U+T | 3D-U型主干 | 3D-SwinTransformer | 卷积层 | 0.003 73 | 35.657 0 | 549.18 |
U+T+DFV | 3D-U型主干 | 3D-SwinTransformer | DFV | 35.660 0 | 549.18 | |
X | 3D-ConvNeXt | — | 卷积层 | 0.005 10 | 9.760 0 | 233.17 |
X+T | 3D-ConvNeXt | 3D-SwinTransformer | 卷积层 | 0.002 21 | 16.251 1 | 248.09 |
X+T+DFV | 3D-ConvNeXt | 3D-SwinTransformer | DFV | 0.000 98 |
模型 | MSE | RMSE | Sqr.rel | Bumpiness | 参数量/103 |
---|---|---|---|---|---|
RDF | 0.111 5 | 0.322 | 0.239 5 | 1.54 | — |
DDFF | 0.033 4 | 0.167 | 1.74 | 39 806 222 | |
DefocusNet | 0.021 8 | 0.134 | 0.035 9 | 2.52 | 1 508 047 |
AiFDepthNet | — | — | 16 533 873 | ||
FV-Net | 0.018 8 | 0.125 | 0.024 3 | 1.45 | |
DFV-Net | 0.020 5 | 0.129 | 0.023 9 | — | |
GSTFC | 0.009 8 | 0.091 | 0.057 5 | 0.51 | 16 269 231 |
表2 不同模型在FoD500数据集上的对比结果
Tab. 2 Comparison results of different models on FoD500 dataset
模型 | MSE | RMSE | Sqr.rel | Bumpiness | 参数量/103 |
---|---|---|---|---|---|
RDF | 0.111 5 | 0.322 | 0.239 5 | 1.54 | — |
DDFF | 0.033 4 | 0.167 | 1.74 | 39 806 222 | |
DefocusNet | 0.021 8 | 0.134 | 0.035 9 | 2.52 | 1 508 047 |
AiFDepthNet | — | — | 16 533 873 | ||
FV-Net | 0.018 8 | 0.125 | 0.024 3 | 1.45 | |
DFV-Net | 0.020 5 | 0.129 | 0.023 9 | — | |
GSTFC | 0.009 8 | 0.091 | 0.057 5 | 0.51 | 16 269 231 |
数据集 | 模型 | RMSE | PSNR/dB | SSIM | Correlation |
---|---|---|---|---|---|
Base-Line | SF | 0.032 6 | 30.100 6 | 0.947 5 | 0.946 5 |
TENV | 0.020 8 | 33.889 6 | 0.935 4 | 0.994 4 | |
DLAP | 0.025 3 | 32.900 6 | 0.910 2 | 0.996 0 | |
FDC | 0.092 4 | 20.875 7 | 0.649 6 | 0.495 4 | |
GC | 0.122 0 | 18.653 7 | 0.676 8 | 0.779 6 | |
RDF | 0.024 7 | 32.847 8 | 0.904 8 | 0.962 4 | |
RFVR-SFF | 0.008 0 | 42.813 1 | 0.952 3 | 0.996 9 | |
GSTFC | 0.005 8 | 45.714 2 | 0.967 6 | 0.998 2 | |
4D Light Field | SF | 0.040 7 | 28.630 6 | 0.899 0 | 0.904 2 |
TENV | 0.042 9 | 28.397 2 | 0.887 5 | 0.882 5 | |
DLAP | 0.030 9 | 31.297 5 | 0.920 8 | 0.945 6 | |
FDC | 0.088 8 | 21.203 1 | 0.563 8 | 0.523 9 | |
GC | 0.112 9 | 19.162 5 | 0.796 0 | 0.697 4 | |
RDF | 0.055 7 | 25.934 5 | 0.871 3 | 0.816 5 | |
RFVR-SFF | 0.029 9 | 32.096 7 | 0.895 0 | 0.934 6 | |
GSTFC | 0.026 2 | 33.686 5 | 0.921 7 | 0.947 6 | |
POV-Ray | SF | 0.099 0 | 20.165 9 | 0.611 7 | 0.731 4 |
TENV | 0.099 8 | 20.249 0 | 0.614 7 | 0.704 3 | |
DLAP | 0.077 1 | 22.361 1 | 0.658 4 | 0.835 9 | |
FDC | 0.123 4 | 18.228 6 | 0.465 4 | 0.508 7 | |
GC | 0.140 3 | 17.086 7 | 0.532 1 | 0.546 2 | |
RDF | 0.112 2 | 19.038 8 | 0.604 4 | 0.666 1 | |
RFVR-SFF | 0.093 6 | 20.802 0 | 0.500 2 | 0.771 8 | |
GSTFC | 0.076 6 | 22.402 4 | 0.655 0 | 0.845 1 | |
SLFD and DLFD | SF | 0.047 8 | 26.754 0 | 0.888 8 | 0.904 5 |
TENV | 0.056 0 | 25.628 3 | 0.857 9 | 0.859 1 | |
DLAP | 0.036 1 | 29.377 6 | 0.919 2 | 0.944 9 | |
FDC | 0.101 4 | 20.026 1 | 0.541 3 | 0.553 6 | |
GC | 0.125 9 | 18.178 3 | 0.776 9 | 0.710 7 | |
RDF | 0.066 1 | 24.223 2 | 0.869 7 | 0.807 4 | |
RFVR-SFF | 0.085 5 | 21.720 0 | 0.388 8 | 0.726 4 | |
GSTFC | 0.038 7 | 28.987 6 | 0.914 5 | 0.927 4 |
表3 不同模型在传统数据集上的对比结果
Tab. 3 Comparison results of different models on traditional datasets
数据集 | 模型 | RMSE | PSNR/dB | SSIM | Correlation |
---|---|---|---|---|---|
Base-Line | SF | 0.032 6 | 30.100 6 | 0.947 5 | 0.946 5 |
TENV | 0.020 8 | 33.889 6 | 0.935 4 | 0.994 4 | |
DLAP | 0.025 3 | 32.900 6 | 0.910 2 | 0.996 0 | |
FDC | 0.092 4 | 20.875 7 | 0.649 6 | 0.495 4 | |
GC | 0.122 0 | 18.653 7 | 0.676 8 | 0.779 6 | |
RDF | 0.024 7 | 32.847 8 | 0.904 8 | 0.962 4 | |
RFVR-SFF | 0.008 0 | 42.813 1 | 0.952 3 | 0.996 9 | |
GSTFC | 0.005 8 | 45.714 2 | 0.967 6 | 0.998 2 | |
4D Light Field | SF | 0.040 7 | 28.630 6 | 0.899 0 | 0.904 2 |
TENV | 0.042 9 | 28.397 2 | 0.887 5 | 0.882 5 | |
DLAP | 0.030 9 | 31.297 5 | 0.920 8 | 0.945 6 | |
FDC | 0.088 8 | 21.203 1 | 0.563 8 | 0.523 9 | |
GC | 0.112 9 | 19.162 5 | 0.796 0 | 0.697 4 | |
RDF | 0.055 7 | 25.934 5 | 0.871 3 | 0.816 5 | |
RFVR-SFF | 0.029 9 | 32.096 7 | 0.895 0 | 0.934 6 | |
GSTFC | 0.026 2 | 33.686 5 | 0.921 7 | 0.947 6 | |
POV-Ray | SF | 0.099 0 | 20.165 9 | 0.611 7 | 0.731 4 |
TENV | 0.099 8 | 20.249 0 | 0.614 7 | 0.704 3 | |
DLAP | 0.077 1 | 22.361 1 | 0.658 4 | 0.835 9 | |
FDC | 0.123 4 | 18.228 6 | 0.465 4 | 0.508 7 | |
GC | 0.140 3 | 17.086 7 | 0.532 1 | 0.546 2 | |
RDF | 0.112 2 | 19.038 8 | 0.604 4 | 0.666 1 | |
RFVR-SFF | 0.093 6 | 20.802 0 | 0.500 2 | 0.771 8 | |
GSTFC | 0.076 6 | 22.402 4 | 0.655 0 | 0.845 1 | |
SLFD and DLFD | SF | 0.047 8 | 26.754 0 | 0.888 8 | 0.904 5 |
TENV | 0.056 0 | 25.628 3 | 0.857 9 | 0.859 1 | |
DLAP | 0.036 1 | 29.377 6 | 0.919 2 | 0.944 9 | |
FDC | 0.101 4 | 20.026 1 | 0.541 3 | 0.553 6 | |
GC | 0.125 9 | 18.178 3 | 0.776 9 | 0.710 7 | |
RDF | 0.066 1 | 24.223 2 | 0.869 7 | 0.807 4 | |
RFVR-SFF | 0.085 5 | 21.720 0 | 0.388 8 | 0.726 4 | |
GSTFC | 0.038 7 | 28.987 6 | 0.914 5 | 0.927 4 |
1 | HUANG B, WANG W Q, BATES M, et al. Three-dimensional super-resolution imaging by stochastic optical reconstruction microscopy[J]. Science, 2008, 319(5864): 810-813. 10.1126/science.1153529 |
2 | 闫涛,钱宇华,李飞江,等. 三维时频变换视角的智能微观三维形貌重建方法[J]. 中国科学:信息科学 (2022-03-27) [2022-05-09].. 10.1360/ssi-2021-0386 |
YAN T, QIAN Y H, LI F J, et al. Intelligent microscopic 3D shape reconstruction method based on 3D time-frequency transformation[J]. SCIENTIA SINICA Informationis (2022-03-27) [2022-05-09].. 10.1360/ssi-2021-0386 | |
3 | 闫涛,陈斌,刘凤娴,等. 基于多景深融合模型的显微三维重建方法[J]. 计算机辅助设计与图形学学报, 2017, 29(9): 1613-1623. 10.3969/j.issn.1003-9775.2017.09.004 |
YAN T, CHEN B, LIU F X, et al. Multi-focus image fusion model for micro 3D reconstruction[J]. Journal of Computer-Aided Design and Computer Graphics, 2017, 29(9): 1613-1623. 10.3969/j.issn.1003-9775.2017.09.004 | |
4 | SCHÖNBERGER J L, FRAHM J M. Structure-from-motion revisited[C]// Proceedings of the 2016 IEEE Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE, 2016: 4104-4113. 10.1109/cvpr.2016.445 |
5 | 武越,苑咏哲,岳铭煜,等. 点云配准中多维度信息融合的特征挖掘方法[J]. 计算机研究与发展, 2022, 59(8): 1732-1741. 10.7544/issn1000-1239.20220042 |
WU Y, YUAN Y Z, YUE M Y, et al. Feature mining method of multi-dimensional information fusion in point cloud registration[J]. Journal of Computer Research and Development, 2022, 59(8): 1732-1741. 10.7544/issn1000-1239.20220042 | |
6 | 吉新新,朴永日,张淼,等. 基于光场焦点堆栈的鲁棒深度估计[J]. 计算机学报, 2022, 45(6): 1226-1240. 10.11897/SP.J.1016.2022.01226 |
JI X X, PIAO Y R, ZHANG M, et al. Robust depth estimation via light field focal stacks[J]. Chinese Journal of Computers, 2022, 45(6): 1226-1240. 10.11897/SP.J.1016.2022.01226 | |
7 | YAN T, WU P, QIAN Y H, et al. Multiscale fusion and aggregation PCNN for 3D shape recovery[J]. Information Sciences, 2020, 536: 277-297. 10.1016/j.ins.2020.05.100 |
8 | ALI U, MAHMOOD M T. Robust focus volume regularization in shape from focus[J]. IEEE Transactions on Image Processing, 2021, 30: 7215-7227. 10.1109/tip.2021.3100268 |
9 | JEON H G, SURH J, IM S, et al. Ring difference filter for fast and noise robust depth from focus[J]. IEEE Transactions on Image Processing, 2020, 29: 1045-1060. 10.1109/tip.2019.2937064 |
10 | HAZIRBAS C, SOYER S G, STAAB M C, et al. Deep depth from focus[C]// Proceedings of the 2018 Asian Conference on Computer Vision, LNCS 11363. Cham: Springer, 2019: 525-541. |
11 | MAXIMOV M, GALIM K, LEAL-TAIXÉ L. Focus on defocus: bridging the synthetic to real domain gap for depth estimation[C]// Proceedings of the 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE, 2020: 1068-1077. 10.1109/cvpr42600.2020.00115 |
12 | WANG N H, WANG R, LIU Y L, et al. Bridging unsupervised and supervised depth from focus via all-in-focus supervision[C]// Proceedings of the 2021 IEEE/CVF International Conference on Computer Vision. Piscataway: IEEE, 2021: 12601-12611. 10.1109/iccv48922.2021.01239 |
13 | YANG F T, HUANG X L, ZHOU Z H. Deep depth from focus with differential focus volume[C]// Proceedings of the 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE, 2022: 12632-12641. 10.1109/cvpr52688.2022.01231 |
14 | LIU Z, NING J, CAO Y, et al. Video Swin Transformer[C]// Proceedings of the 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE, 2022: 3192-3201. 10.1109/cvpr52688.2022.00320 |
15 | NAYAR S K, NAKAGAWA Y. Shape from focus[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 1994, 16(8): 824-831. 10.1109/34.308479 |
16 | PECH-PACHECO J L, CRISTÓBAL G, CHAMORRO-MARTÍNEZ J, et al. Diatom autofocusing in brightfield microscopy: a comparative study[C]// Proceedings of the 15th International Conference on Pattern Recognition. Piscataway: IEEE, 2000: 314-317. |
17 | AN Y, KANG G, KIM I J, et al. Shape from focus through Laplacian using 3D window[C]// Proceedings of the 2nd International Conference on Future Generation Communication and Networking. Piscataway: IEEE, 2008: 46-50. 10.1109/fgcn.2008.139 |
18 | GEUSEBROEK J M, CORNELISSEN F, SMEULDERS A W M, et al. Robust autofocusing in microscopy[J]. Cytometry, 2000, 39(1): 1-9. 10.1002/(sici)1097-0320(20000101)39:1<1::aid-cyto2>3.0.co;2-j |
19 | AHMAD M B, CHOI T S. Application of three dimensional shape from image focus in LCD/TFT displays manufacturing[J]. IEEE Transactions on Consumer Electronics, 2007, 53(1): 1-4. 10.1109/tce.2007.339492 |
20 | HUANG W, JING Z L. Evaluation of focus measures in multi-focus image fusion[J]. Pattern Recognition Letters, 2007, 28(4): 493-500. 10.1016/j.patrec.2006.09.005 |
21 | SUBBARAO M, CHOI T. Accurate recovery of three-dimensional shape from image focus[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 1995, 17(3): 266-274. 10.1109/34.368191 |
22 | CHOI T S, YUN J. Three-dimensional shape recovery from the focused-image surface[J]. Optical Engineering, 2000, 39(5): 1321-1326. 10.1117/1.602498 |
23 | LEE I, MAHMOOD M T, CHOI T S. Adaptive window selection for 3D shape recovery from image focus[J]. Optics and Laser Technology, 2013, 45: 21-31. 10.1016/j.optlastec.2012.08.003 |
24 | THELEN A, FREY S, HIRSCH S, et al. Improvements in shape-from-focus for holographic reconstructions with regard to focus operators, neighborhood-size, and height value interpolation[J]. IEEE Transactions on Image Processing, 2009, 18(1): 151-157. 10.1109/tip.2008.2007049 |
25 | MINHAS R, MOHAMMED A A, WU Q M J. Shape from focus using fast discrete curvelet transform[J]. Pattern Recognition, 2011, 44(4): 839-853. 10.1016/j.patcog.2010.10.015 |
26 | YAN T, HU Z G, QIAN Y H, et al. 3D shape reconstruction from multifocus image fusion using a multidirectional modified Laplacian operator[J]. Pattern Recognition, 2020, 98: No.107065. 10.1016/j.patcog.2019.107065 |
27 | MAHMOOD M T. MRT letter: guided filtering of image focus volume for 3D shape recovery of microscopic objects[J]. Microscopy Research and Technique, 2014, 77(12): 959-963. 10.1002/jemt.22438 |
28 | RIBAL C, LERMÉ N, LE HÉGARAT-MASCLE S. Efficient graph cut optimization for shape from focus[J]. Journal of Visual Communication and Image Representation, 2018, 55: 529-539. 10.1016/j.jvcir.2018.06.029 |
29 | PERTUZ S, PUIG D, GARCIA M A. Analysis of focus measure operators for shape-from-focus[J]. Pattern Recognition, 2013, 46(5): 1415-1432. 10.1016/j.patcog.2012.11.011 |
30 | 山西大学. 一种全局时空聚焦特征耦合的多景深三维形貌重建方法: 202211130317.8[P]. 2022-12-09. |
Shanxi University. A global spatio-temporal focusing feature coupled multi-depth-of-field 3D shape reconstruction method: 202211130317.8[P]. 2022-12-09. | |
31 | SCHECHNER Y Y, KIRYATI N. Depth from defocus vs. stereo: how different really are they?[J]. International Journal of Computer Vision, 2000, 39(2): 141-162. 10.1023/a:1008175127327 |
32 | MUHAMMAD M, CHOI T S. Sampling for shape from focus in optical microscopy[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2012, 34(3): 564-573. 10.1109/tpami.2011.144 |
33 | RONNEBERGER O, FISCHER P, BROX T. U-net: convolutional networks for biomedical image segmentation[C]// Proceedings of the 2015 International Conference on Medical Image Computing and Computer-Assisted Intervention, LNCS 9351. Cham: Springer, 2015: 234-241. |
34 | 孙颖,丁卫平,黄嘉爽,等. RCAR-UNet:基于粗糙注意力机制的视网膜血管分割网络[J/OL]. 计算机研究与发展 (2022-07-13) [2022-09-11].. |
SUN Y, DING W P, HUANG J S, et al. RCAR-UNet: retinal vessels segmentation network based on rough channel attention mechanism[J/OL]. Journal of Computer Research and Development (2022-07-13) [2022-10-19].. | |
35 | LIU Z, MAO H Z, WU C Y, et al. A ConvNet for the 2020s[C]// Proceedings of the 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE, 2022: 11966-11976. 10.1109/cvpr52688.2022.01167 |
36 | 韩冰,张鑫云,任爽. 基于3D点云的卷积运算综述[J/OL]. 计算机研究与发展 (2022-07-11) [2022-11-04].. |
HAN B, ZHANG X Y, REN S. Survey of convolution operations based on 3D point clouds[J/OL]. Journal of Computer Research and Development (2022-07-11) [2022-10-25].. | |
37 | NASEER M, RANASINGHE K, KHAN S H, et al. Intriguing properties of vision transformers[C]// Proceedings of the 35th Conference on Neural Information Processing Systems [2020-10-23].. |
38 | LIU Z, LIN Y T, CAO Y, et al. Swin Transformer: hierarchical vision transformer using shifted windows[C]// Proceedings of the 2021 IEEE/CVF International Conference on Computer Vision. Piscataway: IEEE, 2021: 9992-10002. 10.1109/iccv48922.2021.00986 |
39 | XIE S N, GIRSHICK R, DOLLÁR P, et al. Aggregated residual transformations for deep neural networks[C]// Proceedings of the 2017 IEEE Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE, 2017: 5987-5995. 10.1109/cvpr.2017.634 |
40 | SANDLER M, HOWARD A, ZHU M L, et al. MobileNetV2: inverted residuals and linear bottlenecks[C]// Proceedings of the 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE, 2018: 4510-4520. 10.1109/cvpr.2018.00474 |
41 | SHI J L, JIANG X R, GUILLEMOT C. A framework for learning depth from a flexible subset of dense and sparse light field views[J]. IEEE Transactions on Image Processing, 2019, 28(12): 5867-5880. 10.1109/tip.2019.2923323 |
42 | PERTUZ S, PUIG D, GARCIA M A. Reliability measure for shape-from-focus[J]. Image and Vision Computing, 2013, 31(10): 725-734. 10.1016/j.imavis.2013.07.005 |
43 | HONAUER K, JOHANNSEN O, KONDERMANN D, et al. A dataset and evaluation methodology for depth estimation on 4D light fields[C]// Proceedings of the 2016 Asian Conference on Computer Vision, LNCS 10113. Cham: Springer, 2017: 19-34. |
44 | HEBER S, POCK T. Convolutional networks for shape from light field[C]// Proceedings of the 2016 IEEE Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE, 2016: 3746-3754. 10.1109/cvpr.2016.407 |
[1] | 陈容均, 严宣辉, 杨超城. 面向时间序列的混合图像化循环胶囊分类网络[J]. 《计算机应用》唯一官方网站, 2023, 43(3): 692-699. |
[2] | 伏博毅, 彭云聪, 蓝鑫, 秦小林. 基于深度学习的标签噪声学习算法综述[J]. 《计算机应用》唯一官方网站, 2023, 43(3): 674-684. |
[3] | 王奇, 雷航, 王旭鹏. 姿态干扰下的深度人脸验证[J]. 《计算机应用》唯一官方网站, 2023, 43(2): 595-600. |
[4] | 王萍, 陈楠, 鲁磊. 基于场景先验及注意力引导的跌倒检测算法[J]. 《计算机应用》唯一官方网站, 2023, 43(2): 529-535. |
[5] | 朱利安, 张鸿. 基于双分支条件生成对抗网络的非均匀图像去雾[J]. 《计算机应用》唯一官方网站, 2023, 43(2): 567-574. |
[6] | 沈万里, 张玉金, 胡万. 面向图像修复取证的U型特征金字塔网络[J]. 《计算机应用》唯一官方网站, 2023, 43(2): 545-551. |
[7] | 郭克友, 李雪, 杨民. 基于轻量化YOLOv4的交通信息实时检测方法[J]. 《计算机应用》唯一官方网站, 2023, 43(1): 74-80. |
[8] | 林洋平, 刘佳, 陈培, 张明书, 杨晓元. 基于深度卷积生成对抗网络的半生成式视频隐写方案[J]. 《计算机应用》唯一官方网站, 2023, 43(1): 169-175. |
[9] | 张军, 吴朋莉, 石陆魁, 史进, 潘斌. 联合MOD11A1和地面气象站点数据的多站点温度预测深度学习模型[J]. 《计算机应用》唯一官方网站, 2023, 43(1): 321-328. |
[10] | 王佑芯, 陈斌. 基于深度对比网络的印刷缺陷检测方法[J]. 《计算机应用》唯一官方网站, 2023, 43(1): 250-258. |
[11] | 申志军, 穆丽娜, 高静, 史远航, 刘志强. 细粒度图像分类综述[J]. 《计算机应用》唯一官方网站, 2023, 43(1): 51-60. |
[12] | 李敬虎, 邢前国, 郑向阳, 李琳, 王丽丽. 基于深度学习的无人机影像夜光藻赤潮提取方法[J]. 《计算机应用》唯一官方网站, 2022, 42(9): 2969-2974. |
[13] | 魏佳璇, 杜世康, 于志轩, 张瑞生. 图像分类中的白盒对抗攻击技术综述[J]. 《计算机应用》唯一官方网站, 2022, 42(9): 2732-2741. |
[14] | 尹靖涵, 瞿绍军, 姚泽楷, 胡玄烨, 秦晓雨, 华璞靖. 基于YOLOv5的雾霾天气下交通标志识别模型[J]. 《计算机应用》唯一官方网站, 2022, 42(9): 2876-2884. |
[15] | 张显杰, 张之明. 基于卷积神经网络和Transformer的手写体英文文本识别[J]. 《计算机应用》唯一官方网站, 2022, 42(8): 2394-2400. |
阅读次数 | ||||||
全文 |
|
|||||
摘要 |
|
|||||