Due to the low cost and easy deployment, the depth estimation of monocular cameras has always attractedattention of researchers. As good performance based on deep learning technology in depth estimation, moreand more training models has emerged for depth estimation. Most existing works have required very promisingresults that belongs to supervised learning methods, but corresponding ground truth depth data for training isinevitable that makes training complicated. To overcome this limitation, an unsupervised learning framework isused for monocular depth estimation from videos, which contains depth map and pose network. In this paper,better results can be achieved by optimizing training models and improving training loss. Besides, training andevaluation data is based on standard dataset KITTI(Karlsruhe Institute of Technology and Toyota Institute ofTechnology). In the end, the results are shown through comparing with di erent training models used in thispaper.
展开▼
