Multimodal Image Classification
Multimodal image classification extends standard visual classification by incorporating additional modalities — such as text captions, audio, or structured metadata — alongside image features. Separate encoders process each modality, their representations are fused, and a joint classifier assigns the target label. Models such as CLIP demonstrate that image–text alignment enables zero-shot and few-shot image classification at scale.
Source record
Citations copied verbatim from the method’s source record. No claim-level verification is inferred from them.
- Radford, A., Kim, J. W., Hallacy, C., Ramesh, A., Goh, G., Agarwal, S., ... & Sutskever, I. (2021). Learning transferable visual models from natural language supervision. Proceedings of the 38th International Conference on Machine Learning (ICML), PMLR 139, 8748–8763. · URL
- Ngiam, J., Khosla, A., Kim, M., Nam, J., Lee, H., & Ng, A. Y. (2011). Multimodal deep learning. Proceedings of the 28th International Conference on Machine Learning (ICML), 689–696. · URL
Curated claims
Claims persisted in the evidence ledger, each with its own assessment.
This view does not invent a claim assessment when the ledger has none.
Related methods
Generated from the method graph and shown as machine-suggested relations — no evidence claim is inferred.