Pith. sign in

REVIEW 4 major objections 6 minor 35 references

A smartphone assistant for blind users reports top acceptance scores in a seven-person pilot, with haptic guidance doing the heavy lifting.

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

T0 review · deepseek-v4-flash

2026-08-03 23:20 UTC pith:HGK7HGGV

load-bearing objection Useful assistive-app pilot with a genuinely nice haptic guidance idea, but the satisfaction claim rests on a TAM questionnaire that asks about meetings and classrooms rather than the tasks tested. the 4 major comments →

arxiv 2511.06080 v4 pith:HGK7HGGV submitted 2025-11-08 cs.CV cs.CYcs.HC

AIDEN: Design and Pilot Study of an AI Assistant for the Visually Impaired

Luis Marquez-Carpintero , Francisco Gomez-Donoso , Zuria Bauer , Bessie Dominguez-Dager , Alvaro Belmonte-Baeza , M\'onica Pina-Navarro , Francisco Morillas-Espejo , Felix Escalona
show 1 more author
Miguel Cazorla
This is my paper
classification cs.CV cs.CYcs.HC
keywords assistive technologyvisual impairmenthaptic feedbackobject detectionYOLOLLaVAtechnology acceptance modelmobile accessibility
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

AIDEN is a smartphone assistant that aims to give visually impaired people greater independence by combining real-time object detection, text reading, and scene description in one app. The central claim is that moving spatial guidance out of the auditory channel and into touch—using a Geiger-counter-style vibration that speeds up as the target object approaches the camera frame's center—makes the system intuitive and effective. Pilot results with seven visually impaired participants show high perceived usefulness (4.25/5) and ease of use (4.10/5), with the 'Find an Object' module running at 1.96 frames per second. The authors take this as evidence that multimodal haptic-visual feedback can improve daily usability compared with audio-centric assistants, and as grounds for larger clinical validation.

Core claim

The paper demonstrates, at pilot scale, that a phone-based assistant can unify object search, OCR, and scene understanding through a distributed architecture in which heavy models run on a server while the phone handles capture and feedback. The strongest quantitative claims are the high acceptance scores on the Technology Acceptance Model and the real-time object-finding latency (about 0.5 seconds on the phone), with the qualitative highlight being the Geiger-counter haptic feedback that lets users center objects without listening to continuous audio. The authors also report that Attitude Toward Using, not Perceived Usefulness, best predicts intention to use, suggesting that the experience

What carries the argument

The Geiger-counter haptic guidance loop: the Euclidean distance between the target object's bounding-box center and the camera frame center is inversely mapped to vibration and audio pulse frequency—intermittent when the object is peripheral, rapid as it nears center, and continuous with a confirmation tone when locked. This mechanism is what lets the system offload spatial feedback from the auditory channel to the tactile channel, which the paper presents as its central answer to the auditory-masking problem in existing assistive technology.

Load-bearing premise

The TAM questionnaire includes items about meetings, classrooms, and hybrid teaching that were never part of the evaluated tasks; the high acceptance scores assume those off-target items measure the same usefulness as the actual object-finding, OCR, and scene-description tasks.

What would settle it

Ask a new group of users to rate the three tested tasks separately from the classroom and meeting items; if task-specific usefulness scores drop below the reported 4.25, the central acceptance claim is overstated. Alternatively, run the object-finding task in a noisy real-world environment and compare vibration-only guidance against audio-only guidance; if users center objects equally quickly with audio, the haptic channel's claimed advantage is not supported.

Watch this falsifier. Get emailed when new claim-graph text bears on it.

If this is right

  • Haptic object-centering can work on ordinary smartphones and returns spatial guidance at 1.96 frames per second, fast enough for real-time camera aiming.
  • Users rated the system highly on usefulness, ease of use, attitude, and intention to use; Cronbach's alpha values of 0.82 and 0.93 suggest internally consistent questionnaire scales.
  • Adoption intent correlates most strongly with Attitude Toward Using, not Perceived Usefulness, implying that design quality and enjoyment are decisive for assistive-technology acceptance.
  • The phone-server split keeps heavy AI off the device, making the assistant viable on mid-range hardware, while the stated no-storage policy addresses privacy concerns.
  • The findings support moving to larger, longitudinal, in-the-wild studies to test the system beyond a controlled lab.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The classroom and meeting items in the questionnaire look like they were reused from a different study context; a fair reader should not treat those items as evidence about object-finding or OCR usability, since those scenarios were never tested.
  • If the haptic channel proves robust in noisy outdoor settings, the same Geiger metaphor could be extended to navigation obstacles or door-finding, not just object centering.
  • The reported real-time performance depends on a server connection; on weaker networks the latency could degrade, making an edge/offline version the natural stress test of the claim.
  • A direct head-to-head comparing vibration-only guidance against audio-only guidance on centering speed and accuracy would separate whether haptics, rather than novelty, drive the high satisfaction scores.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit.

Referee Report

4 major / 6 minor

Summary. The manuscript describes AIDEN, a smartphone-based assistive system for visually impaired users combining YOLO object detection, LLaVA-based scene description/OCR, and a Geiger-counter-like haptic guidance mode. The authors report a pilot user study with seven visually impaired participants who performed object retrieval, OCR, and scene-description tasks, followed by a 16-item Technology Acceptance Model questionnaire, and runtime measurements of the three core functions. The central claims are that AIDEN achieves high user satisfaction, particularly in intuitiveness and perceived autonomy, and that the Find-an-Object mode achieves effective real-time performance. The paper also discusses architectural choices, privacy, and limitations.

Significance. If the pilot evidence were adequate, the system would be a useful contribution to assistive technology: it integrates off-the-shelf models into an accessible smartphone interface, uses the haptic channel to avoid auditory masking, and the authors are explicit about hallucination and WCAG limitations. The engineering choices are sensible, and the ONCE-recruited sample, while small, is appropriate for early usability work. However, the evaluation as reported does not support the headline claims: the TAM instrument measures contexts (meetings, classrooms, hybrid teaching) that were not part of the evaluated tasks; there is no baseline or control; task success data are not reported; and the real-time performance label is not justified for 1.96 fps. The paper is best viewed as a design description with preliminary usability observations, not as a validation of acceptance or performance.

major comments (4)
  1. [§4.3, Table 4, §6.2] The TAM instrument has a construct-validity problem. Q5–Q7 and Q12 ask about 'conducting meetings', 'visualizing remote participants in a classroom', 'person visualization system', and 'hybrid teaching sessions'. These contexts appear nowhere in the evaluated tasks (Task A: object retrieval; Task B: OCR; Task C: scene understanding/VQA). Table 4 nevertheless groups Q5–Q7/Q12 as PU1–PU4, and §6.1 defines PU as 'the degree to which the user believes that AIDEN enhances their autonomy in daily tasks'. The reported PU aggregate mean (4.25) and its CI therefore do not measure usefulness for the tasks AIDEN actually supports; they may reflect a generic teaching/meeting system. Since the abstract's 'high user satisfaction' claim is based on this aggregate, the central empirical claim is unsupported. The authors should either administer task-matched items or remove the mismatched items and re-an
  2. [§4.5.2, §6, §8] No control or baseline condition is included, and the task success criterion is defined but the success rates are never reported. The abstract and §7 conclude that multimodal haptic–visual feedback can 'improve daily usability and independence compared to traditional audio-centric methods', but no audio-centric or no-assistant condition was tested. The study is a single-arm pilot in a controlled lab. This is acknowledged as a limitation in §8, but the conclusion in the abstract overreaches. At minimum, the paper should state that no comparative claim is being made, or add a baseline condition/scoring task completion.
  3. [Table 3, §6, §7] The label 'effective real-time performance' for 'Find an Object' is not supported by the reported 1.96 fps (mean smartphone-side time 0.51 s per request). First, 'frames per second' implies a continuous processing rate, but the measurement is a single-request latency; the protocol (20 executions, server/phone times) does not establish an operating frame rate during continuous camera scanning. Second, no rationale is given for why 0.5 s per update is sufficient for the Geiger-counter guidance loop, which updates haptic pulses as the user pans the camera. Standard video-rate processing is far above 1.96 fps, and a 0.5 s latency can be perceptible during motion. The real-time claim should be reworded to 'low-latency per request' or supported with an update-rate requirement and a measurement of continuous-mode throughput.
  4. [§6.2, §7] The inferential statistics are over-interpreted for N=7. The Spearman correlation reported as β=0.869 (p<0.05) between ATT and BI is a single bivariate correlation on seven participants; with such a small sample, the estimate is highly unstable and the p-value is not informative. Similarly, Cronbach's α=0.82/0.93 computed on seven respondents provides only weak evidence of internal consistency, and the Wilcoxon test with p=1.00 for the PU–PEOU comparison is likely an artifact of ties/small sample. These analyses should be clearly labeled exploratory; the discussion's claim that 'ATT appears to be the primary factor driving the intention to use AIDEN' goes beyond what the data can support.
minor comments (6)
  1. [Abstract, §6, §7] The terms 'real-time', 'near real-time', and 'effective real-time' are used inconsistently. Choose one definition and apply it throughout.
  2. [§4] Typo: 'machien learning' should be 'machine learning'.
  3. [Table 1] The note about replacing custom LaTeX commands (\cmark, \xmark, \imark) should be removed before publication.
  4. [Figure 5] Figure 5 is referenced in the text but no image is visible in the manuscript. Ensure the figure is actually included.
  5. [§4.3 / Table 4] Q12 is described in the text as addressing 'hybrid teaching sessions', but Table 4 labels PU4 as 'Satisfies specific need'. The item wording and table label must be aligned.
  6. [§5] The qualitative labels 'Excellent' and 'Best' are borrowed from Bangor et al.'s SUS adjective scale and applied to 5-point Likert TAM items without justification. Either justify this mapping or remove the adjectives.

Circularity Check

0 steps flagged

No circularity: system evaluation uses external models, standard TAM instrument, and measured runtime data; the pilot study does not derive its conclusions from its inputs by construction.

full rationale

The paper reports an empirical pilot study of a smartphone-based assistive system. The central claims — high user satisfaction and real-time performance — are supported by measured runtime data (Table 3) and a standard TAM questionnaire (Table 4), not by a derivation that reintroduces its own assumptions. The system components (YOLO, LLaVA, haptic feedback timings) are external or design parameters, not fitted to the outcome measures. The TAM questionnaire's content validity is questionable: items Q5–Q7 and Q12 ask about meetings, classrooms, and hybrid teaching contexts that were not among the evaluated tasks, and Section 6.1 defines PU as 'the degree to which the user believes that AIDEN enhances their autonomy in daily tasks (e.g., reading without assistance)' while the PU items operationalize meeting/classroom usefulness. This is a measurement-validity concern, not a circularity concern: the reported means are averages of participant responses, not constructed from the claims they support. The sample-size justification, the bootstrapped confidence intervals, and the Wilcoxon test are statistical procedures rather than circular argumentation. The limitation section explicitly states that N=7 limits generalizability and that results are a preliminary validation. No equation or parameter is defined in terms of the dependent variables, and no load-bearing claim is justified solely by a self-citation. Therefore no significant circularity is present.

Axiom & Free-Parameter Ledger

2 free parameters · 5 axioms · 0 invented entities

The central claim of high user satisfaction depends on pre-trained vision models (YOLO/LLaVA) and a standard survey instrument (TAM). The only hand-set parameters are haptic timings and distance thresholds, which are not fitted to data. No new theoretical entities are introduced.

free parameters (2)
  • Haptic pulse timing constants = 200ms ON / 300ms OFF; 200ms ON / 100ms OFF; continuous ON
    Hand-chosen interaction parameters; no justification or user-specific tuning reported.
  • Distance-to-frequency mapping thresholds = not specified
    The mapping between object-center distance d and pulse frequency is described only qualitatively; exact thresholds are not reported, making the haptic algorithm under-specified.
axioms (5)
  • domain assumption YOLOv8 provides reliable detection for the 80 COCO classes plus a custom 'door' class.
    The system depends on YOLOv8's pre-trained weights; no accuracy validation on the target environment or user scenarios is provided.
  • domain assumption LLaVA provides accurate scene description and OCR.
    The system uses LLaVA with fixed prompts; no quantitative OCR accuracy or scene-description quality metrics are reported.
  • domain assumption The operating system's screen reader (TalkBack/VoiceOver) works as expected with the app.
    The UI relies on the OS accessibility tree; no explicit testing of screen reader behavior or compatibility is described.
  • domain assumption The 'five-users' heuristic justifies a sample size of N=7 for usability evaluation.
    Cited to support sample size; it addresses detection of usability problems, not statistical power for acceptance measurement.
  • domain assumption The Technology Acceptance Model (TAM) is a valid framework for assistive technology acceptance.
    TAM is borrowed from information systems research and applied here via prior extensions; construct validity for this specific assistive context is not re-established.

pith-pipeline@v1.3.0-alltime-deepseek · 12161 in / 8903 out tokens · 83342 ms · 2026-08-03T23:20:15.879757+00:00 · methodology

0 comments
read the original abstract

This paper presents AIDEN, an artificial intelligence-based assistant designed to enhance the autonomy and daily quality of life of visually impaired individuals, who often struggle with object identification, text reading, and navigation in unfamiliar environments. Existing solutions such as screen readers or audio-based assistants facilitate access to information but frequently lead to auditory overload and raise privacy concerns in open environments. AIDEN addresses these limitations with a hybrid architecture that integrates You Only Look Once (YOLO) for real-time object detection and a Large Language and Vision Assistant (LLaVA) for scene description and Optical Character Recognition (OCR). A key novelty of the system is a continuous haptic guidance mechanism based on a Geiger-counter metaphor, which supports object centering without occupying the auditory channel, while privacy is preserved by ensuring that no personal data are stored. Empirical evaluations with visually impaired participants assessed perceived ease of use and acceptance using the Technology Acceptance Model (TAM). Results indicate high user satisfaction, particularly regarding intuitiveness and perceived autonomy. Moreover, the ``Find an Object'' achieved effective real-time performance. These findings provide promising evidence that multimodal haptic-visual feedback can improve daily usability and independence compared to traditional audio-centric methods, motivating larger-scale clinical validations.

Figures

Figures reproduced from arXiv: 2511.06080 by Alvaro Belmonte-Baeza, Bessie Dominguez-Dager, Felix Escalona, Francisco Gomez-Donoso, Francisco Morillas-Espejo, Luis Marquez-Carpintero, Miguel Cazorla, M\'onica Pina-Navarro, Zuria Bauer.

Figure 1
Figure 1. Figure 1: User interface views of the application. [PITH_FULL_IMAGE:figures/full_fig_p007_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Random samples of input images and retrieved text returned by the OCR feature. [PITH_FULL_IMAGE:figures/full_fig_p009_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Actual detections of the “Find an Object” method in an office environment. [PITH_FULL_IMAGE:figures/full_fig_p009_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Examples of input images and corresponding answers returned by the Scene Descrip [PITH_FULL_IMAGE:figures/full_fig_p011_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: Average score and standard deviation per question obtained by the 16-item question [PITH_FULL_IMAGE:figures/full_fig_p012_5.png] view at source ↗

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

35 extracted references · 1 linked inside Pith

  1. [1]

    Mobile vision as assistive technology for the blind: An experimental study

    Roberto Manduchi. “Mobile vision as assistive technology for the blind: An experimental study”. In:International Conference on Computers for Handicapped Persons. Springer. 2012, pp. 9–16

    2012

  2. [2]

    Accessed: 2025-11-30

    Seeing AI.Seeing AI - Talking Camera App for the Blind. Accessed: 2025-11-30. 2024. url:https://www.seeingai.com/

    2025

  3. [3]

    Accessed: 2025-11-30

    Be My Eyes.Be My Eyes: Lend Your Eyes to the Blind. Accessed: 2025-11-30. 2024.url: https://www.bemyeyes.com

    2025

  4. [4]

    Enhancing screen reader intelligibility in noisy environments

    Dragan Ahmetovic et al. “Enhancing screen reader intelligibility in noisy environments”. In:IEEE Transactions on Human-Machine Systems53.4 (2023), pp. 771–780

    2023

  5. [5]

    Hearing the World: A Pilot Study Design on Spatial Audio for the Visually Impaired

    Xinyan Zhao. “Hearing the World: A Pilot Study Design on Spatial Audio for the Visually Impaired”. In:Proceedings of the 27th International Academic Mindtrek Conference. 2024, pp. 244–248

    2024

  6. [6]

    Developing apps for people with sensory disabilities, and implications for technology acceptance models

    Paraskevi Theodorou and Apostolos Meliones. “Developing apps for people with sensory disabilities, and implications for technology acceptance models”. In:Global Journal of Information Technology: Emerging Technologies9.2 (2019), pp. 33–40

    2019

  7. [7]

    Evaluation and comparison of artificial intelligence vision aids: Orcam myeye 1 and seeing ai

    Christina Granquist et al. “Evaluation and comparison of artificial intelligence vision aids: Orcam myeye 1 and seeing ai”. In:Journal of Visual Impairment & Blindness115.4 (2021), pp. 277–285

    2021

  8. [8]

    Privacy concerns and behaviors of people with visual impairments

    Tousif Ahmed et al. “Privacy concerns and behaviors of people with visual impairments”. In:Proceedings of the 33rd annual ACM conference on human factors in computing sys- tems. 2015, pp. 3523–3532

    2015

  9. [9]

    Geiger counters

    Serge A Korff. “Geiger counters”. In:Nuclear Instrumentation II/Instrumentelle Hilfs- mittel der Kernphysik II. Springer, 1958, pp. 52–85

    1958

  10. [10]

    Enhancing High-tech Assistive Technology Use for Learning among Students with Visual Impairment in Tanzania’s Higher Education

    Sarah Ezekiel Kisanga. “Enhancing High-tech Assistive Technology Use for Learning among Students with Visual Impairment in Tanzania’s Higher Education”. In:University of Dar es Salaam Library Journal20.1 (2025), pp. 226–239

    2025

  11. [11]

    Leveraging assistive technology for visually im- paired people through optimal deep transfer learning based object detection model

    Mahir Mohammed Sharif Adam et al. “Leveraging assistive technology for visually im- paired people through optimal deep transfer learning based object detection model”. In: Scientific Reports15.1 (2025), p. 30113

    2025

  12. [12]

    Accessed: 2024-09-23

    Mediate.Supersense: AI scanner for the blind.https://www.supersense.app/. Accessed: 2024-09-23. 2024

    2024

  13. [13]

    Factors influencing the intention of persons with visual impair- ment to adopt mobile applications based on the UTAUT model

    Hyunchang Moon et al. “Factors influencing the intention of persons with visual impair- ment to adopt mobile applications based on the UTAUT model”. In:Universal Access in the Information Society21.1 (2022), pp. 93–107

    2022

  14. [14]

    2025.doi:https://arxiv.org/abs/2506.05369

    Nicolas Pfitzer et al.MR.NA VI: Mixed-Reality Navigation Assistant for the Visually Im- paired. 2025.doi:https://arxiv.org/abs/2506.05369. arXiv:2506.05369 [cs.CV]

    Pith/arXiv arXiv 2025

  15. [15]

    Auditory perceptual consolidation in early-onset blindness

    Alexander A Stevens and Kurt Weaver. “Auditory perceptual consolidation in early-onset blindness”. In:Neuropsychologia43.13 (2005), pp. 1901–1910

    2005

  16. [16]

    Accessed: 2025-11-30

    Meta Platforms, Inc.Ray-Ban Meta Smart Glasses.https://www.meta.com/smart- glasses/. Accessed: 2025-11-30. 2024

    2025

  17. [17]

    Technology acceptance model

    Fred D Davis, Richard P Bagozzi, and Paul R Warshaw. “Technology acceptance model”. In:J Manag Sci35.8 (1989), pp. 982–1003

    1989

  18. [18]

    Web Accessibility for Visually Impaired Users: Extending the Technology Acceptance Model (TAM)

    Soussan Djamasbi et al. “Web Accessibility for Visually Impaired Users: Extending the Technology Acceptance Model (TAM)”. In:Proceedings of AMCIS 2006. 2006. 17

    2006

  19. [19]

    Acceptance factors for the use of video call via smartphone by blind people

    Anima Bag et al. “Acceptance factors for the use of video call via smartphone by blind people”. In:Technology in Society(2016)

    2016

  20. [20]

    Challenges in Ac- ceptance of Smartphone-Based Assistive Technologies: Extending the UTAUT Model for People With Visual Impairments

    Paraskevi Theodorou, Kleomenis Tsiligkos, and Apostolos Meliones. “Challenges in Ac- ceptance of Smartphone-Based Assistive Technologies: Extending the UTAUT Model for People With Visual Impairments”. In:Journal of Visual Impairment & Blindness118.1 (2024), pp. 18–30

    2024

  21. [21]

    Accessed: 2023-09-24

    Ionic.Screen Reader Capacitor Plugin API.https://capacitorjs.com/docs/apis/ screen-reader. Accessed: 2023-09-24

    2023

  22. [22]

    Accessed: 2023-09-24

    Kazuya Kawaguchi.Vue I18n.https://vue-i18n.intlify.dev/. Accessed: 2023-09-24

    2023

  23. [23]

    Assistive technology in resource-limited environments: a scoping review

    Rebecca Matter et al. “Assistive technology in resource-limited environments: a scoping review”. In:Disability and Rehabilitation: Assistive Technology12.2 (2017), pp. 105–114

    2017

  24. [24]

    Refining the test phase of usability evaluation: how many subjects is enough?

    Robert A Virzi. “Refining the test phase of usability evaluation: how many subjects is enough?” In:Human factors34.4 (1992), pp. 457–468

    1992

  25. [25]

    A mathematical model of the finding of usabil- ity problems

    Jakob Nielsen and Thomas K Landauer. “A mathematical model of the finding of usabil- ity problems”. In:Proceedings of the INTERACT’93 and CHI’93 conference on Human factors in computing systems. 1993, pp. 206–213

    1993

  26. [26]

    W3C Rec- ommendation

    Andrew Kirkpatrick et al.Web Content Accessibility Guidelines (WCAG) 2.1. W3C Rec- ommendation. World Wide Web Consortium (W3C), June 2018.url:https://www.w3. org/TR/WCAG21/

    2018

  27. [27]

    Version 8.0.0

    Glenn Jocher, Ayush Chaurasia, and Jing Qiu.Ultralytics YOLO. Version 8.0.0. Jan. 2023. url:https://github.com/ultralytics/ultralytics

    2023

  28. [28]

    Protocol analysis

    K Anders Ericsson. “Protocol analysis”. In:A companion to cognitive science(2017), pp. 425–432

    2017

  29. [29]

    Likert scale: Explored and explained

    Ankur Joshi et al. “Likert scale: Explored and explained”. In:British journal of applied science & technology7.4 (2015), p. 396

    2015

  30. [30]

    Determining what individual SUS scores mean: Adding an adjective rating scale

    Aaron Bangor, Philip Kortum, and James Miller. “Determining what individual SUS scores mean: Adding an adjective rating scale”. In:Journal of usability studies4.3 (2009), pp. 114–123

    2009

  31. [31]

    Multivariate data analysis

    Joseph F Hair. “Multivariate data analysis”. In: (2009)

    2009

  32. [32]

    Wilcoxon signed-rank test

    Robert F Woolson. “Wilcoxon signed-rank test”. In:Wiley encyclopedia of clinical trials (2007), pp. 1–3

    2007

  33. [33]

    Examining an extended technology acceptance model with ex- perience construct on hotel consumers’ adoption of mobile applications

    Yu-Chih Huang et al. “Examining an extended technology acceptance model with ex- perience construct on hotel consumers’ adoption of mobile applications”. In:Journal of Hospitality Marketing & Management28.8 (2019), pp. 957–980

    2019

  34. [34]

    Consumer acceptance and use of information technology: extending the unified theory of acceptance and use of technology

    Viswanath Venkatesh, James YL Thong, and Xin Xu. “Consumer acceptance and use of information technology: extending the unified theory of acceptance and use of technology”. In:MIS quarterly(2012), pp. 157–178

    2012

  35. [35]

    Adoption of mobile payment among visually impaired users in Tamil Nadu based on technology acceptance model (TAM)

    V Parvathy and D Durairaj. “Adoption of mobile payment among visually impaired users in Tamil Nadu based on technology acceptance model (TAM)”. In:International journal of health sciences6.S3 (2022), pp. 5346–5361. 18

    2022