Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Stereo Correspondence and Reconstruction of Endoscopic Data Challenge

Max Allan, Jonathan Mcleod, Congcong Wang, Jean Claude Rosenthal, Zhenglei Hu, Niklas Gard, Peter Eisert, Ke Xue Fu, Trevor Zeffiro, Wenyao Xia, Zhanshi Zhu, Huoling Luo, Fucang Jia, Xiran Zhang, Xiaohong Li, Lalith Sharan, Tom Kurmann, Sebastian Schmid, Raphael Sznitman, Dimitris Psychogyios, Mahdi Azizian, Danail Stoyanov, Lena Maier-Hein, Stefanie Speidel

TL;DR

The paper introduces the SCARED endoscopic stereo dataset and Benchmark from MICCAI 2019, addressing the gap in dense depth estimation for surgical scenes. It surveys ten participating methods, spanning traditional stereo pipelines, multi-view cost volumes, and learning-based disparity networks, across seven training and two test sequences with structured-light ground truth. The work highlights the practical challenges of dataset calibration, synchronization, and ground-truth alignment, and reports competition results with Rediminds’ method achieving top performance. It also provides post-challenge methods and a supplemental discussion on dataset inaccuracies to guide future improvements and benchmarking in surgical depth estimation.

Abstract

The stereo correspondence and reconstruction of endoscopic data sub-challenge was organized during the Endovis challenge at MICCAI 2019 in Shenzhen, China. The task was to perform dense depth estimation using 7 training datasets and 2 test sets of structured light data captured using porcine cadavers. These were provided by a team at Intuitive Surgical. 10 teams participated in the challenge day. This paper contains 3 additional methods which were submitted after the challenge finished as well as a supplemental section from these teams on issues they found with the dataset.

Stereo Correspondence and Reconstruction of Endoscopic Data Challenge

TL;DR

The paper introduces the SCARED endoscopic stereo dataset and Benchmark from MICCAI 2019, addressing the gap in dense depth estimation for surgical scenes. It surveys ten participating methods, spanning traditional stereo pipelines, multi-view cost volumes, and learning-based disparity networks, across seven training and two test sequences with structured-light ground truth. The work highlights the practical challenges of dataset calibration, synchronization, and ground-truth alignment, and reports competition results with Rediminds’ method achieving top performance. It also provides post-challenge methods and a supplemental discussion on dataset inaccuracies to guide future improvements and benchmarking in surgical depth estimation.

Abstract

The stereo correspondence and reconstruction of endoscopic data sub-challenge was organized during the Endovis challenge at MICCAI 2019 in Shenzhen, China. The task was to perform dense depth estimation using 7 training datasets and 2 test sets of structured light data captured using porcine cadavers. These were provided by a team at Intuitive Surgical. 10 teams participated in the challenge day. This paper contains 3 additional methods which were submitted after the challenge finished as well as a supplemental section from these teams on issues they found with the dataset.

Paper Structure

This paper contains 27 sections, 8 figures, 3 tables.

Table of Contents

  1. Introduction
  2. Data
  3. Participating Methods
  4. Fraunhofer Institute for Telecommunications, Heinrich Hertz Institute
  5. University of Western Ontario
  6. Norwegian University of Science and Technology
  7. Rediminds Inc.
  8. Unknown
  9. Harbin Institute of Technology
  10. Shenzhen Institute for Advanced Technology (1)
  11. Johns Hopkins University
  12. Shenzhen Institute for Advanced Technology (2)
  13. Heidelberg University
  14. Post-Challenge Methods
  15. Wellcome/EPSRC Centre for Interventional and Surgical Sciences (WEISS) UCL (1)
  16. ...and 12 more sections

Figures (8)

  • Figure 1: An example image captured by the endoscope and the corresponding depth map.
  • Figure 2: (a) The projector used in this data collection. (b) The view of the endoscope while the pattern is being projected. (c) A single example of a Gray code pattern used to create the dataset.
  • Figure 3: Mean absolute error plots for each frame of test dataset 1, keyframes 1 and 2.
  • Figure 4: Mean per-pixel error plots for each frame of test dataset 1, keyframes 3 and 4.
  • Figure 5: Mean per-pixel error plots for each frame of test dataset 2, keyframes 1 and 2.
  • ...and 3 more figures