This deep learning model leverages CLIP (Contrastive Language-Image Pretraining) to classify images based on user-defined class names or descriptions. By utilizing CLIP’s extensive pre-trained understanding of visual and textual relationships, the model can accurately categorize images according to the input classes. It supports classification of individual images or batches of images from a folder and is well-suited for oriented imagery. Alongside the assigned class label, the model also outputs a probability score indicating the confidence of the classification.