Cutup and Detect: Human Fall Detection on Cutup Untrimmed Videos Using a Large Foundational Video Understanding Model

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Grutschus, Till, Karrar, Ola, Esenov, Emir, Vats, Ekta
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866914657472610304
author Grutschus, Till
Karrar, Ola
Esenov, Emir
Vats, Ekta
author_facet Grutschus, Till
Karrar, Ola
Esenov, Emir
Vats, Ekta
contents This work explores the performance of a large video understanding foundation model on the downstream task of human fall detection on untrimmed video and leverages a pretrained vision transformer for multi-class action detection, with classes: "Fall", "Lying" and "Other/Activities of daily living (ADL)". A method for temporal action localization that relies on a simple cutup of untrimmed videos is demonstrated. The methodology includes a preprocessing pipeline that converts datasets with timestamp action annotations into labeled datasets of short action clips. Simple and effective clip-sampling strategies are introduced. The effectiveness of the proposed method has been empirically evaluated on the publicly available High-Quality Fall Simulation Dataset (HQFSD). The experimental results validate the performance of the proposed pipeline. The results are promising for real-time application, and the falls are detected on video level with a state-of-the-art 0.96 F1 score on the HQFSD dataset under the given experimental settings. The source code will be made available on GitHub.
format Preprint
id arxiv_https___arxiv_org_abs_2401_16280
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Cutup and Detect: Human Fall Detection on Cutup Untrimmed Videos Using a Large Foundational Video Understanding Model
Grutschus, Till
Karrar, Ola
Esenov, Emir
Vats, Ekta
Computer Vision and Pattern Recognition
This work explores the performance of a large video understanding foundation model on the downstream task of human fall detection on untrimmed video and leverages a pretrained vision transformer for multi-class action detection, with classes: "Fall", "Lying" and "Other/Activities of daily living (ADL)". A method for temporal action localization that relies on a simple cutup of untrimmed videos is demonstrated. The methodology includes a preprocessing pipeline that converts datasets with timestamp action annotations into labeled datasets of short action clips. Simple and effective clip-sampling strategies are introduced. The effectiveness of the proposed method has been empirically evaluated on the publicly available High-Quality Fall Simulation Dataset (HQFSD). The experimental results validate the performance of the proposed pipeline. The results are promising for real-time application, and the falls are detected on video level with a state-of-the-art 0.96 F1 score on the HQFSD dataset under the given experimental settings. The source code will be made available on GitHub.
title Cutup and Detect: Human Fall Detection on Cutup Untrimmed Videos Using a Large Foundational Video Understanding Model
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2401.16280