Jun’21: EndoSLAM Dataset is now published at Mendeley!

09 Jun 2021

Our dataset associated to the paper "EndoSLAM dataset and an unsupervised monocular visual odometry and depth estimation approach for endoscopic videos" is publlished at Mendeley Data. 


EndoSLAM dataset includes standard and capsule endoscopy data from ex-vivo porcine gastrointestinal tract organs with time-synchronized, high-precision 6D ground trruth pose, as well as a video sequence recorded by a clinically in-use colonoscope from a fully representative silicon colon phantom with CT ground truth. Besides, it includes synthetically generated data with pixel-wise depth and 6D pose ground truth which facilitates the simulation to real domain adaptation algorithms.