Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Saraiva, Pedro Antonio Rabelo, de Souza, Enzo Ferreira, Pinheiro, Joao Manoel Herrera, Segreto, Thiago H., Godoy, Ricardo V., Becker, Marcelo
Format: Preprint
Veröffentlicht: 2025
Schlagworte:
Online-Zugang:https://arxiv.org/abs/2508.17468
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866912772275568640
author Saraiva, Pedro Antonio Rabelo
de Souza, Enzo Ferreira
Pinheiro, Joao Manoel Herrera
Segreto, Thiago H.
Godoy, Ricardo V.
Becker, Marcelo
author_facet Saraiva, Pedro Antonio Rabelo
de Souza, Enzo Ferreira
Pinheiro, Joao Manoel Herrera
Segreto, Thiago H.
Godoy, Ricardo V.
Becker, Marcelo
contents This paper addresses the challenges of data scarcity and high acquisition costs in training robust object detection models for complex industrial environments, such as offshore oil platforms. Data collection in these hazardous settings often limits the development of autonomous inspection systems. To mitigate this issue, we propose a hybrid data synthesis pipeline that integrates procedural rendering and AI-driven video generation. The approach uses BlenderProc to produce photorealistic images with domain randomization and NVIDIA's Cosmos-Predict2 to generate physically consistent video sequences with temporal variation. A YOLO-based detector trained on a composite dataset, combining real and synthetic data, outperformed models trained solely on real images. A 1:1 ratio between real and synthetic samples achieved the highest accuracy. The results demonstrate that synthetic data generation is a viable, cost-effective, and safe strategy for developing reliable perception systems in safety-critical and resource-constrained industrial applications.
format Preprint
id arxiv_https___arxiv_org_abs_2508_17468
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle A Synthetic Dataset for Manometry Recognition in Robotic Applications
Saraiva, Pedro Antonio Rabelo
de Souza, Enzo Ferreira
Pinheiro, Joao Manoel Herrera
Segreto, Thiago H.
Godoy, Ricardo V.
Becker, Marcelo
Computer Vision and Pattern Recognition
Artificial Intelligence
Machine Learning
Robotics
This paper addresses the challenges of data scarcity and high acquisition costs in training robust object detection models for complex industrial environments, such as offshore oil platforms. Data collection in these hazardous settings often limits the development of autonomous inspection systems. To mitigate this issue, we propose a hybrid data synthesis pipeline that integrates procedural rendering and AI-driven video generation. The approach uses BlenderProc to produce photorealistic images with domain randomization and NVIDIA's Cosmos-Predict2 to generate physically consistent video sequences with temporal variation. A YOLO-based detector trained on a composite dataset, combining real and synthetic data, outperformed models trained solely on real images. A 1:1 ratio between real and synthetic samples achieved the highest accuracy. The results demonstrate that synthetic data generation is a viable, cost-effective, and safe strategy for developing reliable perception systems in safety-critical and resource-constrained industrial applications.
title A Synthetic Dataset for Manometry Recognition in Robotic Applications
topic Computer Vision and Pattern Recognition
Artificial Intelligence
Machine Learning
Robotics
url https://arxiv.org/abs/2508.17468