Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Concept Videos

Depth Perception and Spatial Vision01:15

Depth Perception and Spatial Vision

Depth perception is the ability to perceive objects three-dimensionally. It relies on two types of cues: binocular and monocular. Binocular cues depend on the combination of images from both eyes and how the eyes work together. Since the eyes are in slightly different positions, each eye captures a slightly different image. This disparity between images, known as binocular disparity, helps the brain interpret depth. When the brain compares these images, it determines the distance to an object.

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same journal

Machine Learning-Based Privacy Preserving via CT/MRI and Organ Metadata Prediction.

Journal of imaging informatics in medicine·2026
Same journal

AI-Based Segmental Volumetry of the Downstream Aorta in Aortic Dissection: End-to-End Versus Hybrid Strategies.

Journal of imaging informatics in medicine·2026
Same journal

BrainGAN Framework for Generating Synthetic Contrast-Enhanced Multisequence MRI in Brain Metastases.

Journal of imaging informatics in medicine·2026
Same journal

Anatomically Localized Detection of Six Acute Abdominal Emergencies on CT Using Multi-window Deep Learning: Development and Validation.

Journal of imaging informatics in medicine·2026
Same journal

DSGH-Net: Medical Image Segmentation via Dual-Statistical Dynamic Context and Graph-Convolutional Heterogeneous Decoder.

Journal of imaging informatics in medicine·2026
Same journal

Acceptance and Utilization of Picture Archiving and Communication Systems in Emergency Departments: A TAM-Based Study in Palestinian Governmental Hospitals.

Journal of imaging informatics in medicine·2026

Related Experiment Video

Updated: Jun 20, 2026

Robotized Testing of Camera Positions to Determine Ideal Configuration for Stereo 3D Visualization of Open-Heart Surgery
05:12

Robotized Testing of Camera Positions to Determine Ideal Configuration for Stereo 3D Visualization of Open-Heart Surgery

Published on: August 12, 2021

Self-supervised Depth Estimation for Monocular Endoscopy Using Confidence-Rectified Distillation and Semantic

Zhongping Li1, Kejin Zhu1, Guozhe Jin2

  • 1College of Engineering, Yanbian University, Yanji, Jilin Province, China.

Journal of Imaging Informatics in Medicine
|June 18, 2026
PubMed
Summary
This summary is machine-generated.

This study introduces confidence-rectified distillation and semantic distribution alignment to improve monocular depth estimation (MDE) in endoscopy. These methods enhance accuracy and cross-domain generalization for 3D reconstruction and augmented reality applications.

Keywords:
Confidence-rectified distillationMonocular depth estimationMulti-resolution prediction consistency filterSelf-supervised learningSemantic distribution alignment

More Related Videos

End-To-End Deep Neural Network for Salient Object Detection in Complex Environments
03:31

End-To-End Deep Neural Network for Salient Object Detection in Complex Environments

Published on: December 15, 2023

Related Experiment Videos

Last Updated: Jun 20, 2026

Robotized Testing of Camera Positions to Determine Ideal Configuration for Stereo 3D Visualization of Open-Heart Surgery
05:12

Robotized Testing of Camera Positions to Determine Ideal Configuration for Stereo 3D Visualization of Open-Heart Surgery

Published on: August 12, 2021

End-To-End Deep Neural Network for Salient Object Detection in Complex Environments
03:31

End-To-End Deep Neural Network for Salient Object Detection in Complex Environments

Published on: December 15, 2023

Area of Science:

  • Computer Vision
  • Medical Imaging
  • 3D Reconstruction

Background:

  • Monocular depth estimation (MDE) is crucial for 3D reconstruction and augmented reality in endoscopy.
  • Existing self-supervised methods face challenges due to specular highlights, low texture, and non-Lambertian effects in endoscopic images, degrading depth accuracy.
  • There is a need for improved MDE techniques that are robust to these challenging visual conditions and generalize across different endoscopic datasets.

Purpose of the Study:

  • To enhance the accuracy and cross-domain generalization of self-supervised monocular depth estimation in endoscopic imaging.
  • To introduce novel methods that address the specific challenges posed by endoscopic visual characteristics like specularities and low texture.
  • To improve the robustness and reliability of depth estimation for applications in 3D reconstruction and intraoperative augmented reality.

Main Methods:

  • Developed confidence-rectified distillation using a Multi-Resolution Prediction Consistency Filter (MRPCF) to estimate per-pixel confidence and weight supervision.
  • Implemented semantic distribution alignment (SDA) by aligning semantic features from a frozen DINOv3 teacher to the student's geometric space.
  • Utilized global log-domain alignment of multi-resolution pseudo-labels to improve robustness to specular outliers and scale variations.

Main Results:

  • The proposed approach significantly surpasses existing self-supervised methods on the official SCARED dataset.
  • Demonstrated consistent performance gains on the Hamlyn Heart and SERV-CT datasets in zero-shot evaluations, indicating strong cross-dataset generalization.
  • The methods effectively improve boundary fidelity and preserve thin structures by matching semantic distributions.

Conclusions:

  • The combined approach of confidence-rectified distillation and semantic distribution alignment offers a robust solution for monocular depth estimation in challenging endoscopic environments.
  • The proposed techniques enhance depth accuracy and generalization capabilities, paving the way for more reliable 3D reconstruction and augmented reality in surgery.
  • Future work can explore further refinements of these methods for real-time applications and diverse medical imaging modalities.