Student or educator? Sign up with your school email and get 6 months of Pro free.
CoChat

The DOI matches its Crossref record; 39 of the 44 references it cites were found in a registry or an index; 2 of the 14 sources its claims cite were checked for support.

StereoPilot: A Wearable Target Location System for Blind and Visually Impaired Using Spatial Audio Rendering

Open access
Hu, Xuhui, Song, Aiguo, Wei, Zhikai, Zeng, Hong · 2022 · IEEE Transactions on Neural Systems and Rehabilitation Engineering

Abstract

Vision loss severely impacts object recognition and spatial cognition for limited vision individuals. It is a challenge to compensate for this using other sensory modalities, such as touch or hearing. This paper introduces StereoPilot, a wearable target location system to facilitate the spatial cognition of BVI. Through wearing a head-mounted RGB-D camera, the 3D spatial information of the environment is measured and processed into navigation cues. Leveraging spatial audio rendering (SAR) technology, it allows the navigation cues to be transmitted in a type of 3D sound from which the sound orientation can be distinguished by the sound localization instincts in humans. Three haptic and auditory display strategies were compared with SAR through experiments with three BVI and four sighted subjects. Compared with mainstream speech instructional feedback, the experimental results of the Fitts' law test showed that SAR increases the information transfer rate (ITR) by a factor of three for spatial navigation, while the positioning error is reduced by 40%. Furthermore, SAR has a lower learning effect than other sonification approaches such as vOICe. In desktop manipulation experiments, StereoPilot was able to obtain precise localization of desktop objects while reducing the completion time of target grasping tasks in half as compared to the voice instruction method. In summary, StereoPilot provides an innovative wearable target location solution that swiftly and intuitively transmits environmental information to BVI individuals in the real world.

Overview

CoChat Summary

TL;DR

StereoPilot is a wearable target-location system that uses a head-mounted RGB-D camera and spatial audio rendering (SAR) to convey object locations to blind and visually impaired users. In experiments with 3 BVI and 4 sighted participants, SAR improved information transfer rate and reduced positioning error and grasping time relative to voice, vibrotactile, and vOICe-style sonification feedback.

Key findings

  • In the Fitts’ law test, BVI participants had SAR ITR of 0.764 bits/s versus VI 0.243, VB 0.256, and NS −0.596; completion time was 9.97 ± 5.71 s for SAR versus 22.46 ± 9.12 s for VI, 14.01 ± 7.42 s for VB, and 33.04 ± 19.57 s for NS.
  • For BVI participants, SAR reduced positioning error by 28% compared with VI, 39% compared with VB, and 46% compared with NS; SAR positioning accuracy for BVI was 22% higher than for sighted subjects. The abstract reports SAR increased ITR by a factor of three and reduced positioning error by 40% compared with speech instruction.
  • In desktop manipulation, the RGB-D camera’s average localization error was 0.43 ± 0.14 cm; the system ran at 15 fps with 43% CPU and 400 MB memory usage. SAR shortened target-grasping completion time, reported in the abstract as half that of the voice instruction method.
  • Success rates were similar across SAR, VI, and VB but lower for NS: BVI SAR 96.61 ± 1.43%, VI 95.90 ± 1.06%, VB 96.24 ± 1.51%, NS 84.56 ± 3.54%; sighted SAR 97.11 ± 1.82%, VI 95.59 ± 1.93%, VB 95.60 ± 1.07%, NS 87.61 ± 4.60%.

Methods

  • The wearable system used an Intel RealSense D435 RGB-D camera and MPU9250 on a head-mounted helmet, MediaPipe for object and hand recognition, and coordinate transformation from the camera coordinate system to the user coordinate system; stereo earphones provided audio feedback.
  • SAR encoded target location using a sound cue with high- and low-frequency components, a CIPIC-based HRTF model, and Google Resonance Audio in a Unity virtual scene, with the virtual hand as listener and target object as sound source.
  • Experiments included 4 sighted participants (25.5 ± 4.2 years) and 3 participants with congenital blindness (22.6 ± 1.5 years), all with normal hearing. The Fitts’ law test compared SAR, VI, VB, and NS with 30 repetitions per group; the desktop manipulation experiment used five same-material, same-mass, same-shape blocks differing in color, randomly positioned less than 5 cm apart, with 30 trials per feedback method.

Limitations

  • Stated: some subjects reported that rendered spatial audio deviated from actual orientation, possibly due to the generalized HRTF model; the authors propose subject-specific HRTF and more realistic rendering.
  • Stated: machine vision recognition errors during human-environment interaction, especially hand/object occlusion, can cause positioning errors and incorrect spatial information; robustness needs improvement.
  • Stated: vOICe/NS had a 1 fps snapshot update rate, high learning cost, and target/cursor confusion; BVI success was lower than sighted partly because RGB-D recognition accuracy decreased when the hand blocked the object. Evident: the sample was small (3 BVI, 4 sighted), limiting generalizability.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.1109/tnsre.2022.3182661

Cites

39 of 39 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Figures

Fig. 1 The design concept of StereoPilot.
Fig. 1 The design concept of StereoPilot.CC-BY · From the PDF
Fig. 2 (a)The system framework and (b)the typical workflow of StereoPilot. (TTS: text to speech; SAR: spatial audio rendering)
Fig. 2 (a)The system framework and (b)the typical workflow of StereoPilot. (TTS: text to speech; SAR: spatial audio rendering)CC-BY · From the PDF
Fig. 3 The virtual environment for spatial audio rendering. (a) Audio room, (b) virtual hand, and (c) target object with virtual sound source.
Fig. 3 The virtual environment for spatial audio rendering. (a) Audio room, (b) virtual hand, and (c) target object with virtual sound source.CC-BY · From the PDF
Fig. 4 The experimental setup of the desktop manipulation experiment.
Fig. 4 The experimental setup of the desktop manipulation experiment.CC-BY · From the PDF
Fig. 5 The user interface of Fitts' law test.
Fig. 5 The user interface of Fitts' law test.CC-BY · From the PDF
Fig. 6 Scatter plot based on MT and ID and the linear regression curve. For simplicity, only a portion of the sample points are shown.
Fig. 6 Scatter plot based on MT and ID and the linear regression curve. For simplicity, only a portion of the sample points are shown.CC-BY · From the PDF
Fig. 7 (a) Positioning error of BVI and sighted subjects in Fitts' law test. (b) Learning effect of four spatial information feedback methods.
Fig. 7 (a) Positioning error of BVI and sighted subjects in Fitts' law test. (b) Learning effect of four spatial information feedback methods.CC-BY · From the PDF
Fig. 8 3D coordinate positioning accuracy of the RGB-D camera.
Fig. 8 3D coordinate positioning accuracy of the RGB-D camera.CC-BY · From the PDF
Fig. 9 (a) Success rate and (b) completion time in desktop manipulation experiment.
Fig. 9 (a) Success rate and (b) completion time in desktop manipulation experiment.CC-BY · From the PDF

Verification

Claims the paper makes

1 Valid · 5 Not valid · 8 Not checked · 53 Own results and interpretations

35 more claims

Create a free account to see all 67 claims and to verify the 3 references still open.

See all claims — free account
  • Probably not a substantial claim. 16
  • It may state two separate claims as one. 17
  • The quote only partly states this claim. 1

Create a free account to see these claims.

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

Each contradiction between two passages listed here was confirmed by an independent check.

1 of 1

Citations

Cites

Show all 44 references

Cited by

See all 50

Metadata

Authors
Hu, X., Song, A., Wei, Z., & Zeng, H.
Affiliations
State Key Laboratory of Digital Medical Engineering (CN)
Venue
IEEE Transactions on Neural Systems and Rehabilitation Engineering
Published
2022
Volume
30
Pages
1621-1630
ISSN
  • 1534-4320 (ISSN-L)
  • 1558-0210
OpenAlex
W4282935978
Cited by
50
References
44
Citations by year
2026: 11 · 2025: 20 · 2024: 11 · 2023: 7 · 2022: 1
FWCI
5.03
Type
journal article
Language
en
Open access
Locations
  • IEEE Transactions on Neural Systems and Rehabilitation Engineering (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion)
Funders
  • National Natural Science Foundation of China (92148205, 62173089)
  • Basic Research Program of Jiangsu Province (BK201900240)
  • Basic Research Funds for Universities (2242022K30056)
Topics
  • Life Sciences › Neuroscience › Cognitive Neuroscience › Tactile and Sensory Interactions
  • Health Sciences › Health Professions › Occupational Therapy › Assistive Technology in Communication and Mobility
  • Social Sciences › Human Factors and Ergonomics › Digital Accessibility for Disabilities
Keywords
blind and visually impaired, spatial audio rendering, sonification, haptic feedback, spatial cognition, Fitts' law, information transfer rate, object localization, assistive technology, sound localization
MeSH terms
Wearable Electronic Devices, Blindness, Humans, Vision, Ocular, Visual Perception, Persons with Visual Disabilities

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

The claim verdicts this report counts are experimental: their accuracy is not yet measured.

A language model drew the claims from the paper on 2026-10-10, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).

A second model labelled each claim on 2026-10-10 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).

Cited sources are checked with a language model’s help (qwen3.8-omni-flash).

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

How to cite

Cite CoChat for the checks on this page, and the paper for its own findings.

This page
CoChat. (2026). Citation check: StereoPilot: A Wearable Target Location System for Blind and Visually Impaired Using Spatial Audio Rendering. https://app.cochat.ai/papers/10.1109/tnsre.2022.3182661
The paper
Hu, X., Song, A., Wei, Z., & Zeng, H. (2022). StereoPilot: A Wearable Target Location System for Blind and Visually Impaired Using Spatial Audio Rendering. IEEE Transactions on Neural Systems and Rehabilitation Engineering. https://doi.org/10.1109/tnsre.2022.3182661

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.