Search research articles

ABOUT JoVE

Overview Leadership Blog JoVE Help Center

AUTHORS

Publishing Process Editorial Board Scope & Policies Peer Review FAQ Submit

LIBRARIANS

Testimonials Subscriptions Access Resources Library Advisory Board FAQ

RESEARCH

JoVE Journal Methods Collections JoVE Encyclopedia of Experiments Archive

EDUCATION

JoVE Core JoVE Business JoVE Science Education JoVE Lab Manual Faculty Resource Center Faculty Site

Terms & Conditions of Use

Search research articles

Related Experiment Video

Updated: Jan 16, 2026

Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications

Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications

Published on: December 15, 2023

A novel 3D indoor localization method integrating deep spatial feature augmentation and attention-based denoising.

Chunyuan Liu¹, Xiaomin Yu², Peilong Wu¹

¹School of Computer and Control Engineering, Qiqihar University, Heilongjiang, 161006, Qiqihar, China.

Scientific Reports

|September 26, 2025

Summary

Related Concept Videos

Depth Perception and Spatial Vision

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 author

Spider-web-inspired phosphorylated aerogel engineered via Schiff base linkages for efficient uranium capture.

Journal of hazardous materials·2026

Same author

Precision management of the pernicious weed Eriochloa villosa (Thunb.) Kunth in maize: an anchor-point calibration framework integrating economic thresholds and species distribution models.

Pest management science·2026

Same author

An evaluation of the anti-human respiratory syncytial virus activity of sophora alkaloids and the enhanced efficacy by liposomes via intratracheal administration.

Colloids and surfaces. B, Biointerfaces·2026

Same author

A multicenter predictive model for inadvertent intraoperative hypothermia management in elderly patients in Southwest China.

Scientific reports·2026

Same author

Potentiating immunotherapy in "immune-cold" solid tumors through orchestrating T cell immunity via tumor-specific genetic engineering.

Cell reports. Medicine·2025

Same author

Maintenance of long-term efficacy with continuous baricitinib treatment in patients with severe alopecia areata: 3-year results from BRAVE-AA1 and BRAVE-AA2.

Journal of the American Academy of Dermatology·2025

This study introduces a new 3D indoor positioning method using deep learning to improve accuracy with limited data. The approach enhances spatial features and denoises signals, significantly boosting localization performance.

Area of Science:

Computer Science
Artificial Intelligence
Robotics

Background:

Three-dimensional (3D) indoor positioning systems face challenges due to complex environments and high-dimensional data.
Existing methods struggle with scarce training data, noise, and capturing global spatial features, limiting real-world application.
High data acquisition costs and human effort for indoor positioning data collection are significant barriers.

Purpose of the Study:

To develop a novel 3D indoor positioning method that overcomes limitations of data scarcity and poor robustness.
To enhance deep spatial feature extraction and implement attention-based denoising for improved accuracy.
To generate high-quality, high-density 3D positioning data from limited real samples.

Main Methods:

Keywords:

3D indoor positioning Attention mechanism Denoising autoencoder Stacked variational autoencoder Wasserstein generative adversarial network

Related Experiment Videos

Last Updated: Jan 16, 2026

Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications

Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications

Published on: December 15, 2023

Utilized a stacked variational autoencoder (SVAE) for structured deep spatial representation extraction.

Employed a Wasserstein generative adversarial network (WGAN) to synthesize realistic high-density samples, addressing data sparsity.

Integrated an attention mechanism for enhanced global feature perception and controlled noise injection for robustness.

Main Results:

Achieved 100% building and 94.7% floor localization accuracy on the UJIIndoorLoc dataset with only 10% of data.
Reduced positioning error by 14.32% when combined with a deep neural network (DNN).
Demonstrated significant improvements on Tampere and UTSIndoorLoc datasets, with floor localization accuracies of 92.83% and 94.33%, and error reductions of 15.18% and 18.89% respectively.

Conclusions:

The proposed method effectively enhances 3D indoor positioning accuracy and robustness, even with limited training data.
Deep spatial feature enhancement and attention-based denoising are crucial for overcoming current system limitations.
The approach offers a viable solution for cost-effective and accurate indoor positioning systems.