Deep Triple-Supervision Learning Unannotated Surgical Endoscopic Video Data for Monocular Dense Depth Estimation | IEEE Conference Publication | IEEE Xplore