CoChat
Sign Up
Cited by 8Context-Aware Activity Recognition SystemsHuman Pose and Action RecognitionBalance, Gait, and Falls Prevention

PIFR: A novel approach for analyzing pose angle-based human activity to automate fall detection in videos

Open access
Kong, Vungsovanreach, Soeng, Saravit, Thon, Munirot, Cho, Wan-Sup, Nayyar, Anand, Kim, Tae-Kyung·2025·PLOS One
DOI 10.1371/journal.pone.0325253
Open this paper in CoChat — free accountOpen PDF

Abstract

Falls pose a significant health risk for elderly populations, necessitating advanced monitoring technologies. This study introduces a novel two-stage fall detection system that combines computer vision and machine learning to accurately identify fall events. The system uses the YOLOv11 object detection model to track individuals and estimate their body pose by identifying 17 key body points across video frames. The proposed approach extracts nine critical geometric features, including the center of mass and various body angles. These features are used to train a support vector machine (SVM) model for binary classification, distinguishing between standing and lying with high precision. The system’s temporal validation method analyzes sequential frame changes, ensuring robust fall detection. Experimental results, evaluated on the University of Rzeszow Fall Detection (URFD) dataset and the Multiple Cameras Fall Dataset (MCFD), demonstrate exceptional performance, achieving 88.8% precision, 94.1% recall, an F1-score of 91.4%, and a specificity of 95.6%. The method outperforms existing approaches by effectively capturing complex geometric changes during fall events. The system is applicable to smart homes, wearable devices, and healthcare monitoring platforms, offering a scalable, reliable, and efficient solution to enhance safety and independence for elderly individuals, thereby contributing to advancements in health-monitoring technology.

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.1371/journal.pone.0325253

Cites

49 of 49 references verified0 citation errors · 0 not checkable

Statistics

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

  • This performance aligns with recent findings in sensor-based fall detection, where ensemble methods often outperform single classifiers due to their ability to model non-linear biomechanical boundaries without overfitting.

    “This performance aligns with recent findings in sensor-based fall detection, where ensemble methods often outperform single classifiers due to their ability to model non-linear biomechanical boundaries without overfitting [16,20].”

    From Safe Fall: Use of Predictive Modeling and Machine Vision Techniques for Fall Analysis and Fall Quality (2026) · cites this paper as [20]

    The classifier comparison in §5.2 and Table 3 (§5.3.2; pages unavailable) reports that the single SVM outperformed random forest and extreme gradient boosting on video-derived pose features. The paper attributes handling non-linear relationships and resistance to overfitting to SVM, not to ensemble methods in sensor-based fall detection. Although §6.1 says random forest achieved the highest metrics, that conflicts with the reported comparison and does not establish the cited claim. (LLM, conf=0.84)

    Checked by a reasoning model — may be wrong.

    Experimental — accuracy not yet measured

  • Random Forest was selected due to its strong performance in handling tabular kinematic data and its high interpretability compared to more opaque models, as reported in previous sensor-based classification studies.

    “Random Forest was selected due to its strong performance in handling tabular kinematic data and its high interpretability compared to more opaque models, as reported in previous sensor-based classification studies [16,20].”

    From Safe Fall: Use of Predictive Modeling and Machine Vision Techniques for Fall Analysis and Fall Quality (2026) · cites this paper as [20]

    Table 3 (section 5.3.2; page not provided) reports strong Random Forest performance on video-derived pose features, supporting that part of the rationale. The paper does not compare model interpretability or report a sensor-based classification study; section 5.2 says the authors selected SVM for their own classifier. (LLM, conf=0.87)

    Experimental — accuracy not yet measured

  • Kong et al. (2025) achieved an F1-score of 91.4% using an SVM on geometric features.

    “For instance, Wang et al. (2024) reported an accuracy of 89.99% using Random Forest with BlazePose features [14], and Kong et al. (2025) achieved an F1-score of 91.4% using an SVM on geometric features [20].”

    From Safe Fall: Use of Predictive Modeling and Machine Vision Techniques for Fall Analysis and Fall Quality (2026) · cites this paper as [20]

    Section 5.3.1 and Table 4 report a 91.4% F1-score for the overall fall-detection system, which uses pose-angle features and an SVM. But Table 3 reports the SVM classifier’s own F1-score as 98.94%, so 91.4% should not be presented as the SVM’s individual result. (LLM, conf=0.88)

    Experimental — accuracy not yet measured

  • Our proposed framework compares favourably with recent vision-based benchmarks.

    “Our proposed framework compares favourably with recent vision-based benchmarks. For instance, Wang et al. (2024) reported an accuracy of 89.99% using Random Forest with BlazePose features [14], and Kong et al. (2025) achieved an F1-score of 91.4% using an SVM on geometric features [20].”

    From Safe Fall: Use of Predictive Modeling and Machine Vision Techniques for Fall Analysis and Fall Quality (2026) · cites this paper as [20]

Self-check

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

The self-check needs the paper’s full text.

Cites

Flagged first, then by citations

Research works printed but not linked: 2

  1. Powers DMW. Evaluation: From precision, recall and f-measure to roc, informedness, markedness & correlation. arXiv preprint. 2020. https://doi. org/10.48550/arXiv.2010.16061https://arxiv.org/abs/2010.16061. · DOI · Paper page
  2. Asif U, Mashford B, Von Cavallar S, Yohanandan S, Roy S, Tang J. Privacy-preserving human fall detection using video data. In: Machine Learning for Health Workshop, 2020. 39-51.

Cited by

Sorted by citations

Related

Authors
Kong, V., Soeng, S., Thon, M., Cho, W.-S., Nayyar, A., & Kim, T.-K.
Affiliations
  • Chungbuk National University (KR)
  • Duy Tan University (VN)
Corresponding author
Kim, Tae-Kyung
ORCID
Venue
PLOS One
Publisher
Public Library of Science (PLoS)
Published
2025-06-17
Volume
20
Issue
6
Pages
e0325253
ISSN
1932-6203 (ISSN-L)
DOI
10.1371/journal.pone.0325253
OpenAlex
W4411377560
Cited by
8
References
49
Citations by year
2026: 7 · 2025: 1
FWCI
2.79
Type
journal article
Language
en
Open access
Licence
Locations
  • PLoS ONE (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
Funders
Chungbuk National University
Topics
  • Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Context-Aware Activity Recognition Systems
  • Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Human Pose and Action Recognition
  • Health Sciences › Health Professions › Physical Therapy, Sports Therapy and Rehabilitation › Balance, Gait, and Falls Prevention
Keywords
fall detection, human pose estimation, elderly monitoring, YOLOv11 object detection, support vector machine classification, geometric features, temporal validation, smart home safety
MeSH terms
Accidental Falls, Machine Learning, Aged, Algorithms, Humans, Posture, Video Recording, Support Vector Machine

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.

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