Creating good and defective product data necessary for AI visual inspection using 3DCG!
Introduction of AI visual inspection systems. Are you giving up due to a lack of image data for good and defective products?
In recent years, the use of AI in appearance inspection systems has advanced, making it possible to automate complex discrimination and recognition compared to traditional methods. On the other hand, the introduction of AI requires a learning process for the AI model, which necessitates the collection of vast amounts of real image data, presenting a significant barrier to implementation. In particular, collecting images of defective products takes many days, and we have received numerous voices from the field indicating that it is difficult to train the model on defects that have not yet occurred. At our company, leveraging our long-standing experience in 3D design, we can create 3DCG models that serve as alternatives to the learning data for AI models. Notably, we can reproduce various defective product patterns, allowing us to train the AI model in advance on defect patterns that have not yet occurred. *For more details, please refer to the materials. Feel free to contact us with any inquiries.*
- Company：ティーネットジャパン （製造DX支援事業）
- Price：Other
