Elice Group Partners with Hyundai Motor Company to Classify Crash Test Images Using AI

[Financial News] Elice Group announced on the 9th that, in collaboration with Hyundai Motor Company Namyang Research Institute, it has developed an AI solution tailored for manufacturing environments that automatically classifies, searches, and manages crash test images.
The project, led by Hyundai Motor Company Namyang Research Institute, was carried out over approximately eight months, from the end of March to the end of November this year.
Hyundai Motor Company Namyang Research Institute initiated this project to address the challenge of test photos being stored randomly, which required significant manpower and time for classification and retrieval. Elice Group developed a deep learning-based image classification and search model, integrating various modules into a unified service. Notably, the model was trained within a Private Cloud environment using Elice’s modular security system, reducing the risk of data leakage and enhancing security.
The classification of crash test images, which was difficult to achieve with commercial models alone, was resolved by applying a customized AI model retrained on official crash test data. As a result, the system can now classify over 60 types of test images with more than 98% accuracy. Furthermore, by retraining the AI with public datasets, the model was optimized for the test environment at the Namyang Research Institute, even when data was limited.
The image search function allows users to select damaged areas or specific parts in a photo with a mouse, and then displays similar images in order. This feature enables more precise retrieval of comparable test cases than traditional text or image search methods.
Elice Group has integrated image indexing, automatic classification, and search functions into a single desktop program, allowing the entire process from image storage to utilization to be handled automatically.
Jae Won Kim, CEO of Elice Group, stated, “The Hyundai Motor Company Namyang Research Institute, which participated in this project, gave highly positive feedback on the image similarity search function and the program’s user interface. Building on this collaboration, we plan to continuously expand AI solutions that can be immediately applied to real-world tasks in manufacturing, mobility, and various other industries.”
mkchang@fnnews.com Jang Min-kwon Reporter