An Improved 3D Skeletons UP-Fall Dataset: Enhancing Data Quality for Efficient Impact Fall Detection

2025-02-26

AI Summary

Existing UP-Fall datasets struggle to distinguish impact-free sliding motions from genuine ground-impact falls, leading to high false-positive detection rates. To address this, we introduce Impact-Fall: the first high-quality, 3D skeleton-enhanced variant of UP-Fall explicitly designed for impact-event recognition. Our methodology integrates multi-view Kinect v2 acquisition, skeletal keypoint filtering, dynamics-informed impact-timing validation, and a physically grounded, redefined annotation protocol to improve both physical plausibility and inter-class separability. Evaluated on the binary impact-fall classification task, standard models—including LSTM, GCN, and Transformer—achieve an average accuracy gain of 12.7% and an F1-score improvement of 14.3% over baseline UP-Fall. Moreover, models trained on Impact-Fall demonstrate markedly enhanced generalization and robustness across diverse motion patterns and noise conditions. The dataset is publicly released to serve as a more reliable benchmark for fall detection research.

Technology Category

• Intelligent Robots: Multimodal Perception & Sensor Fusion
• Computer Vision: Motion & Tracking
• Knowledge Representation and Reasoning: Geometric, Spatial, and Temporal Reasoning

Application Category

• Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Security and Privacy: Data transparency and provenance

Abstract

Detecting impact where an individual makes contact with the ground within a fall event is crucial in fall detection systems, particularly for elderly care where prompt intervention can prevent serious injuries. The UP-Fall dataset, a key resource in fall detection research, has proven valuable but suffers from limitations in data accuracy and comprehensiveness. These limitations cause confusion in distinguishing between non-impact events, such as sliding, and real falls with impact, where the person actually hits the ground. This confusion compromises the effectiveness of current fall detection systems. This study presents enhancements to the UP-Fall dataset aiming at improving it for impact fall detection by incorporating 3D skeleton data. Our preprocessing techniques ensure high data accuracy and comprehensiveness, enabling a more reliable impact fall detection. Extensive experiments were conducted using various machine learning and deep learning algorithms to benchmark the improved 3D skeletons dataset. The results demonstrate substantial improvements in the performance of fall detection models trained on the enhanced dataset. This contribution aims to enhance the safety and well-being of the elderly population at risk. To support further research and development of building more reliable impact fall detection systems, we have made the improved 3D skeletons UP-Fall dataset publicly available at this link https://zenodo.org/records/12773013.

Problem

• Improving fall detection accuracy
• Enhancing UP-Fall dataset quality
• Distinguishing impact vs. non-impact events

Innovation

• Enhanced 3D skeleton data accuracy
• Improved dataset for fall detection
• Public availability of UP-Fall dataset