Inference for Classifications

Classification models label each sample (or frame) with a Classification, which you generate by passing a model to apply_model().

With a zoo model, this includes zero-shot classifiers like CLIP, which require no training data at all.

Walk through zero-shot classification with a zoo model

For a custom model, you can either construct Classification instances yourself in a loop, or configure a TorchImageModel with a ClassifierOutputProcessor so it works with apply_model() directly, as described in Inference with custom models.

See the manual-loop recipe for custom classifiers