Models for car quality estimation in the
https://github.com/FeMa42/quality_assessment_library repository. It is a simplyfied form of the DamianBoborzi/CarQualityClassifier.
While the CarQualityClassifier models process 4 views as a sequence this model uses 1 view for classifying the model quality. This way this model can be used independent on the amount of views you want to evaluate.
This library provides a set of functions for assessing the quality of images and 3D objects. It is focused on estimating the quality of 3D objects using images.
It provides two main functionalities:
Download the model from
Huggingface and put them in the
models folder. If you want to use the combined model also load the PCA model from the same repository. We provide an example on how to load the models in the
car_quality_metrics.ipynb notebook. Example usage of the car quality metrics can also be found in
car_quality_metrics.ipynb.
