2. 文献情報
K. Tateno, F. Tombari, I. Laina, N. Navab,
"CNN-SLAM: Real-time dense monocular SLAM
with learned depth prediction,“
IEEE Computer Society Conference on
Computer Vision and Pattern Recognition
(CVPR), 2017.
2
21. Related work
Monocular SLAM手法
• DTAM(Dense SLAM)
• LSD-SLAM(Semi-Dense SLAM)
特徴量を利用する手法 画像直接を利用する手法
• ORB-SLAM
ORB特徴量を利用して
疎なSLAMを行う
Raul Mur-Artal, et al., “ORB-SLAM: a Versatile
and Accurate Monocular SLAM System”
Richard A. N., et al., “DTAM:
Dense Tracking and Mapping
in Real-Time”
J.Engel, et.al, “LSD-SLAM:
Large-Scale Direct
Monocular SLAM”
21
29. CNNベースの深度推定と意味付け
Iro Laina, et al.,
“Deeper Depth Prediction with Fully Convolutional Residual Networks”
深度推定
入力:304×228
→
出力:160×128
ResNet-50をImageNetで事前学習したもの
Up Sampling の改良としての Up Projection Layer
29
30. Keisuke Tateno, et.al, “Real-Time and Scalable Incremental
Segmentation on Dense SLAM”
CNNベースの深度推定と意味付け
A. Uckermann, et.al,
“3D scene segmentation for autonomous robot grasping”
を利用した,SLAM用のSegmentationフレームワーク
30
44. 参考文献
• [J. Engel, et al., 2014] J. Engel, T. Schps, and D. Cremers,
“LSD-SLAM: Large-Scale Direct Monocular SLAM.” In European Conference on
Computer Vision (ECCV), 2014.
• [I. Laina, et al., 2016]
I. Laina, C. Rupprecht, V. Belagiannis, F. Tombari, and N. Navab.
“Deeper depth prediction with fully convolutional residual networks.” In IEEE
International Conference on 3D Vision (3DV) (arXiv:1606.00373), October 2016.
• [K. Tateno, et al., 2015]
K. Tateno, F. Tombari, N. Navab, “Real-Time and Scalable Incremental
Segmentation on Dense SLAM”, In IROS. IEEE, 2015.
• [J. Engel, et al., 2013]
J. Engel, J. Sturm, D. Cremers, “Semi-dense visual odometry for a monocular
camera”, In IEEE International Conference on Computer Vision (ICCV), 2013.
44