Сравнение на методи

	Мултимодална класификация на изображения	Мултимодален Трансформер
Област	Дълбоко обучение	Дълбоко обучение
Семейство	Machine learning	Machine learning
Година на възникване	2011–2021	2019–2021
Създател	Ngiam et al.; Radford et al. (CLIP)	Lu et al. (ViLBERT); Radford et al. (CLIP)
Тип	Multimodal supervised classification	Cross-modal attention-based deep learning model
Основополагащ източник	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.	Lu, J., Batra, D., Parikh, D., & Lee, S. (2019). ViLBERT: Pretraining Task-Agnostic Visiolinguistic Representations for Vision-and-Language Tasks. Advances in Neural Information Processing Systems (NeurIPS), 32.
Други названия	multimodal visual classification, image-text classification, vision-language classification, cross-modal image classification	multimodal attention model, cross-modal transformer, vision-language transformer, multi-modal fusion transformer
Свързани	6	5
Резюме	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.	A Multimodal Transformer extends the standard Transformer architecture to process and jointly reason over two or more input modalities — most commonly text and images, but also audio, video, or structured data. Cross-modal attention layers allow information from one modality to inform representations in another, enabling tasks such as visual question answering, image captioning, and multimodal sentiment analysis.