论文笔记总结_S2D_从稀疏到稠密

柔光的暖阳◎ 2022-12-17 11:55 179阅读 0赞

*  1. [2018-ICRA\_稀疏到稠密：从稀疏深度样本+单一图像的深度预测][2018-ICRA]
 *  
 *  以下4篇为传统方法：
 *  2. [2013-CVPR\_结合三维场景重建和类别分割][2013-CVPR]
 *  3. [2010-BMVC\_目标类别分割和稠密立体重建的联合优化][2010-BMVC]
 *  4. [2011-CVPR\_对象立体化-联合立体匹配与对象分割][2011-CVPR_-]
 *  5.[ 2012-ECCV\_从立体图像中提取与场景一致的三维对象和深度][2012-ECCV]
 *  
 *  6. [2018-BMVC\_用于实时语义分割的轻量级精细网络RefineNet][2018-BMVC_RefineNet]
 *  7. [2017-CVPR\_RefineNet: 用于高分辨率语义分割的多路径精细网络][2017-CVPR_RefineNet_]
 *  8. [2019-ICRA\_使用非对称标注，实时联合语义分割和深度估计][2019-ICRA]
 *  9. [2018-ECCV\_从RGB和稀疏感知估计深度][2018-ECCV_RGB]
    
     *  这篇文章内容很接近第一篇
 *  10. [2016-NIPS\_室外单张图像的深度感知][2016-NIPS]
 *  11. [2018-ECCV\_用于语义分割和深度估计的联合任务递归学习][2018-ECCV]
 *  12. [2019-ITSC\_DFuseNet: 用于图像稠密深度补全的、RGB图像和稀疏深度信息深度融合的网络][2019-ITSC_DFuseNet_ _RGB]
    
     *  开源代码：[https://github.com/ShreyasSkandanS/DFuseNet][https_github.com_ShreyasSkandanS_DFuseNet]
 *  13. [2017-3DV-稀疏不变的卷积神经网络（Sparsity Invariant CNNs）][2017-3DV-_Sparsity Invariant CNNs]
    
     *  这篇文章的思路可以参考： 提出一种用于稀疏点输入的CNN，将稀疏点上采样成稠密点
 *  14. [2014-NIPS\_利用多尺度深度网络从单张图像预测深度图][2014-NIPS]
    
     *  这篇文章的思路可以参考
 *  15. [2016-CVPR-ResNet\_用于图像识别的深度残差学习网络][2016-CVPR-ResNet]
 *  16. [2016-3DV-利用全卷积残差网络进行深度预测][2016-3DV-]
    
     *  开源代码：[https://github.com/iro-cp/FCRN-DepthPrediction][https_github.com_iro-cp_FCRN-DepthPrediction]
 *  17. [2018-ECCV-通过卷积空间传播网络的相似性学习进行深度估计][2018-ECCV-]
    
     *  开源代码：[https://github.com/XinJCheng/CSPN][https_github.com_XinJCheng_CSPN]
 *  18. [2019-ICRA\_DeepFusion:基于单视图深度和梯度预测的单目SLAM实时稠密三维重建][2019-ICRA_DeepFusion_SLAM]
 *  19. [2018-PR\_基于膨胀卷积神经网络与软加权和推理的、分层融合单目深度估计][2018-PR]
 *  20. [2017-ICCV-从单张RGB图像到精确尺度深度图评估的一种双支网络][2017-ICCV-_RGB]

[2018-ICRA]: https://blog.csdn.net/shyjhyp11/article/details/107815337
[2013-CVPR]: https://blog.csdn.net/shyjhyp11/article/details/107938439
[2010-BMVC]: https://blog.csdn.net/shyjhyp11/article/details/107944780
[2011-CVPR_-]: https://blog.csdn.net/shyjhyp11/article/details/107961377
[2012-ECCV]: https://blog.csdn.net/shyjhyp11/article/details/107966224
[2018-BMVC_RefineNet]: https://blog.csdn.net/shyjhyp11/article/details/108054612
[2017-CVPR_RefineNet_]: https://blog.csdn.net/shyjhyp11/article/details/108068644
[2019-ICRA]: https://blog.csdn.net/shyjhyp11/article/details/108069338
[2018-ECCV_RGB]: https://blog.csdn.net/shyjhyp11/article/details/108142187
[2016-NIPS]: https://blog.csdn.net/shyjhyp11/article/details/108194916
[2018-ECCV]: https://blog.csdn.net/shyjhyp11/article/details/108398480
[2019-ITSC_DFuseNet_ _RGB]: https://blog.csdn.net/shyjhyp11/article/details/108420508
[https_github.com_ShreyasSkandanS_DFuseNet]: https://github.com/ShreyasSkandanS/DFuseNet
[2017-3DV-_Sparsity Invariant CNNs]: https://blog.csdn.net/shyjhyp11/article/details/108444741
[2014-NIPS]: https://blog.csdn.net/shyjhyp11/article/details/108508460
[2016-CVPR-ResNet]: https://blog.csdn.net/shyjhyp11/article/details/108509410
[2016-3DV-]: https://blog.csdn.net/shyjhyp11/article/details/108509963
[https_github.com_iro-cp_FCRN-DepthPrediction]: https://github.com/iro-cp/FCRN-DepthPrediction
[2018-ECCV-]: https://blog.csdn.net/shyjhyp11/article/details/108579312
[https_github.com_XinJCheng_CSPN]: https://github.com/XinJCheng/CSPN
[2019-ICRA_DeepFusion_SLAM]: https://blog.csdn.net/shyjhyp11/article/details/108821116
[2018-PR]: https://blog.csdn.net/shyjhyp11/article/details/108974323
[2017-ICCV-_RGB]: https://blog.csdn.net/shyjhyp11/article/details/109190009