Using deep learning to find defects.
Deep learning is at the upper end of AI complexity, sifting through more data to achieve more accurate results. Charlie Zhu, vice president of R&D at CyberOptics, talks about how DL can be utilized with inspection to identify defects in chips that are not discernible by traditional computer vision algorithms, classifying multiple objects simultaneously from multiple angles and taking into account reflectivity, variation, and lighting that is not ideal to identify which anomalies will cause serious problems in the field.
 |
 |
 |
 |
 |
 |
 |  |
Leave a Reply