Visual Deformation Detection Using Soft Material Simulation for Pre-training of Condition Assessment Models

TL;DR

This paper presents a method using Blender to generate synthetic datasets for training ML models in deformation detection, improving robustness and reducing reliance on human assessment in manufacturing quality assurance.

Contribution

It introduces a novel pipeline for creating synthetic deformation datasets with customizable parameters and evaluates its impact on ML model performance.

Findings

Synthetic datasets improve model robustness to real-world variations

Background and viewpoint diversity enhance model sensitivity

Pipeline validated with soda can deformation experiment

Abstract

This paper addresses the challenge of geometric quality assurance in manufacturing, particularly when human assessment is required. It proposes using Blender, an open-source simulation tool, to create synthetic datasets for machine learning (ML) models. The process involves translating expert information into shape key parameters to simulate deformations, generating images for both deformed and non-deformed objects. The study explores the impact of discrepancies between real and simulated environments on ML model performance and investigates the effect of different simulation backgrounds on model sensitivity. Additionally, the study aims to enhance the model's robustness to camera positioning by generating datasets with a variety of randomized viewpoints. The entire process, from data synthesis to model training and testing, is implemented using a Python API interfacing with Blender. An experiment with a soda can object validates the accuracy of the proposed pipeline.