MétaCan
Menu
← Back to cohort
Record W7132983618

Advanced AI-based Methods for Automated and Accessible Slip Resistance Evaluation of Winter Footwear

2025· dissertation· W7132983618 on OpenAlexaboutno aff
Shaghayegh Chavoshian

Bibliographic record

VenueTSpace · 2025
Typedissertation
Language
FieldHealth Professions
TopicBalance, Gait, and Falls Prevention
Canadian institutionsnot available
Fundersnot available
KeywordsSlip (aerodynamics)SegmentationBiomechanicsUser interfaceMachine visionAutomation
DOInot available

Abstract

fetched live from OpenAlex

Slips and falls represent a major public health concern in Canada, leading to significant hospitalizations and healthcare costs. Slips are one of the leading causes of falls, which can result in severe injuries, such as strains, sprains, and back injuries. Footwear is a key factor in preventing such incidents, as it serves as the primary interface between the body and walking surfaces. Traditional methods for evaluating footwear slip resistance, such as human-centered and mechanical tests, are typically limited to laboratory settings and are labor-intensive. To address these challenges, this thesis introduces a novel computer vision–driven framework that uses artificial intelligence (AI) to assess and predict footwear slip resistance property. This thesis presents three main contributions: (1) development and prediction of machine learning models that estimate human-centered slip resistance scores using only mechanical test data, enabling manufacturers to estimate human trial outcomes without the need for specialized facilities using a web app. This solution empowers manufacturers to reduce development costs and improve product safety through a data-driven model. (2) Enhancing human-centered testing using computer vision. Our vision transformer model, which is trained on annotated video data, enables real-time and automatic slip detection, eliminating the inconsistency associated with human observers. Furthermore, by integrating participant anthropometrics and gait biomechanics with footwear characteristics, we aim to personalize transformer-based predictive models to estimate slip resistance scores across a more diverse population, thereby making our human-centered tests more generalizable. (3) We introduce a fused model to estimate the slip resistance of winter footwear directly from outsole images. Using image segmentation and convolutional neural networks, our approach identifies tread-ground contact zones and materials to predict slip resistance scores. This end-to-end pipeline is implemented in a mobile application, making safety evaluations accessible to the public and promoting informed footwear choices to reduce winter slip-and-fall injuries. Los resbalones y caídas representan un importante problema de salud pública en Canadá, que conlleva a un número significativo de hospitalizaciones y altos costos para el sistema de salud. Los resbalones son una de las principales causas de caídas, las cuales pueden resultar en lesiones graves, como distensiones, esguinces y lesiones en la espalda. El calzado es un factor clave en la prevención de estos incidentes, ya que constituye la interfaz principal entre el cuerpo y las superficies de marcha. Los métodos tradicionales para evaluar la resistencia al deslizamiento del calzado, como las pruebas centradas en humanos y las pruebas mecánicas, suelen estar limitados a entornos de laboratorio y requieren gran esfuerzo. Para abordar estos desafíos, esta tesis introduce un novedoso marco impulsado por visión por computadora que utiliza inteligencia artificial (IA) para evaluar y predecir la resistencia al deslizamiento del calzado. Esta tesis presenta tres contribuciones principales: (1) Desarrollo y predicción de modelos de aprendizaje automático que estiman puntuaciones de resistencia al deslizamiento centradas en humanos utilizando únicamente datos de pruebas mecánicas, lo que permite a los fabricantes estimar los resultados de ensayos con personas sin necesidad de instalaciones especializadas mediante una aplicación web. Esta solución les permite reducir los costos de desarrollo y mejorar la seguridad del producto mediante un modelo basado en datos. (2) Mejora de las pruebas centradas en humanos mediante visión por computadora. Nuestro modelo de transformador visual, entrenado con datos de video anotados, permite la detección automática y en tiempo real de resbalones, eliminando la inconsistencia asociada a los observadores humanos. Además, al integrar parámetros antropométricos y biomecánica de la marcha de los participantes con las características del calzado, buscamos personalizar los modelos predictivos basados en transformadores para estimar puntuaciones de resistencia al deslizamiento en una población más diversa, haciendo que nuestras pruebas centradas en humanos sean más generalizables. (3) Introducimos un modelo fusionado para estimar la resistencia al deslizamiento del calzado de invierno directamente a partir de imágenes de la suela. Mediante segmentación de imágenes y redes neuronales convolucionales, nuestro enfoque identifica las zonas de contacto suela–suelo y los materiales para predecir puntuaciones de resistencia al deslizamiento. Esta canalización de extremo a extremo se implementa en una aplicación móvil, haciendo que las evaluaciones de seguridad sean accesibles al público y promoviendo elecciones de calzado informadas para reducir las lesiones por resbalones y caídas en invierno.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.

How this classification was reachedexpand

Full frame machine prediction

Teacher imitation

Not calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.

metaresearch head score (Codex)0.001
metaresearch head score (Gemma)0.005
Version: metacan-v3-hybrid-931329e0061cValidation status: machine_predicted_unvalidated
Candidate categoriesnone
Consensus categoriesnone
DomainCandidate signal: none · Consensus signal: none
Study designCandidate signal: Bench or experimental · Consensus signal: none
GenreCandidate signal: Empirical · Consensus signal: none
Teacher disagreement score0.010
Threshold uncertainty score0.020

Distilled classifier scores by category (both heads)

CategoryCodexGemma
Metaresearch0.0010.005
Meta-epidemiology (narrow)0.0010.000
Meta-epidemiology (broad)0.0010.001
Bibliometrics0.0020.001
Science and technology studies0.0000.000
Scholarly communication0.0010.001
Open science0.0010.001
Research integrity0.0010.001
Insufficient payload (model declined to judge)0.0020.001

Machine scores (provisional)

The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.

Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.

Opus teacher head0.080
GPT teacher head0.556
Teacher spread0.475 · how far apart the two teachers sit on this one work
Validation statusscore_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it

Classification

machine, unvalidated

Machine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.

The models applied no category: nothing in the taxonomy fit this work.
Study designBench or experimental
Domainnot available
GenreEmpirical

How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".

Quick stats

Citations0
Published2025
Admission routes1
Has abstractyes

Explore more

Same venueTSpace→Same topicBalance, Gait, and Falls Prevention→French-language works237,207→