OPUS


Volltext-Downloads (blau) und Frontdoor-Views (grau)

Evaluation of 3D Human Pose Estimation Frameworks on Digital Image Data for Forensic Application

Evaluation von 3D Human Pose Estimation Frameworks auf digitalem Bildmaterial für den Einsatz in der Forensik

  • In this thesis a 3D Human Pose Estimation Framework, specifically MMPose using RTM-m Pose and MotionBERT, will be tested. For each created rig anthropometric measurements corresponding to the body parts will be calculated and the rigs compared using the RMSD to review the accuracy and performance regarding various 3D-HPE challenges.

Download full text files

Export metadata

Additional Services

Search Google Scholar

Statistics

frontdoor_oas
Metadaten
Author:Marie Althof
URN:urn:nbn:de:bsz:mit1-opus4-161222
Advisor:Dirk Labudde, Sabine Ute Richter
Document Type:Bachelor Thesis
Language:English
Date of Publication (online):2025/06/12
Year of first Publication:2025
Publishing Institution:Hochschule Mittweida
Granting Institution:Hochschule Mittweida
Date of final exam:2025/04/15
Release Date:2025/06/12
GND Keyword:Pose; Maschinelles Sehen; Objektverfolgung; Bewegungsablauf
Page Number:57
Institutes:Angewandte Computer‐ und Bio­wissen­schaften
DDC classes:363.25 Kriminalistik, Tatrekonstruktion
Open Access:Frei zugänglich