Background: The increasing prevalence of traffic accidents and sports-related injuries highlights the need for advanced tools to predict and analyze these incidents. This project aims to leverage recent advancements in AI-driven motion tracking and finite element (FE) modeling to develop a robust system capable of accurately predicting injury mechanisms in both traffic and sports accidents. Building on previous research (NAISS 2023/22-708, SNIC 2022/22-770), including the development of markerless motion capture systems and personalized human body models, this continuation project will integrate AI with biomechanics to enhance injury prediction and prevention strategies. Introduction: Accident reconstruction using finite element human body models (HBMs) has become an essential tool in understanding and preventing injuries across various scenarios. Despite recent advancements, the field still faces significant challenges, particularly when limited to single-view video sources, which are common in both traffic and sports environments. Accurately replicating the complex dynamics of real-world incidents and the detailed biomechanics of the human body is crucial for precise injury prediction. Advances in finite element modeling, such as the development of baseline male and female pedestrian models within the SAFER HBM framework and methodologies for subject personalization, have marked significant progress. These developments, coupled with advancements in computer vision and AI-driven motion tracking, offer new opportunities to refine FE simulations by incorporating accurate body posture and kinematics from video data. This project seeks to address the following critical issues: 1. Impact of AI-Based Motion Tracking on FE Model Accuracy: Exploring how variations in Skinned Multi-Person Linear (SMPL) estimation methods affect the precision of FE simulations in replicating human body dynamics during accidents. 2. Adaptability Across Accident Scenarios: Enhancing the versatility and applicability of FE simulations by applying these AI-driven methods to diverse accident scenarios, including traffic incidents and sports injuries. Objective: The primary objective of this project is to utilize and enhance the monocular markerless motion capture system developed in prior research, applying it to real-world traffic and sports accident scenarios. This system will analyze video footage to identify the biomechanical parameters involved in injury-causing incidents, providing valuable data for FE simulations. Expected Outcome: Development of an Integrated Analytical Pipeline: A robust pipeline will be developed, capable of processing video data from both traffic and sports scenarios to inform FE simulations with detailed 3D kinematics. Accurate Injury Prediction: By accurately estimating the biomechanical parameters leading to injuries, this project aims to provide a comprehensive understanding of injury mechanisms in traffic and sports accidents. Improved Safety Recommendations: The insights gained will inform the development of better safety measures, equipment, and policies aimed at reducing the risk of injuries in both domains.