• Official

Car Seat Defects Detection

Updated: Oct 22, 2018


This model is capable of dealing with many types of defects on car seat, for example, wrinkles, damage, and dirt etc.




Data Requirements


This model need the user to upload training data associating with labels to use it. In order to train the model, the user can upload as little as 50 images per class.


Performance Metrics


  • Overall Precision: 99.5%

  • Recall: 99%

  • Training Time: estimate 1 hr


Demo Request