Shape-Net: Room Layout Estimation from Panoramic Images Robust to Occlusion using Knowledge Distillation with 3D Shapes as Additional Inputs

no code implementations • 25 Apr 2023 • Mizuki Tabata, Kana Kurata, Junichiro Tamamatsu

It is possible to obtain 3D shape information of rooms as drawings of buildings and coordinates of corners from image datasets, thus we propose providing both 2D panoramic and 3D information to a model to effectively deal with occlusion.

3D Reconstruction Knowledge Distillation +1

Non-learning Stereo-aided Depth Completion under Mis-projection via Selective Stereo Matching

no code implementations • 4 Oct 2022 • Yasuhiro Yao, Ryoichi Ishikawa, Shingo Ando, Kana Kurata, Naoki Ito, Jun Shimamura, Takeshi Oishi

Moreover, under various LiDAR-camera calibration errors, the proposed method reduced the depth estimation MAE to 0. 34-0. 93 times from previous depth completion methods.

Camera Calibration Depth Completion +2