Osadzenia zdań multimodalnych
Osadzenia zdań multimodalnych mapują tekst i obrazy (a czasem dźwięk lub wideo) do wspólnej ciągłej przestrzeni wektorowej, tak aby semantycznie powiązane pary z różnych modalności znalazły się blisko siebie. Trenowane za pomocą celów kontrastywnych na dużych korpusach sparowanych danych, reprezentacje te zasilają wyszukiwanie między modalnościami, klasyfikację zero-shot i rozumowanie wizualno-językowe.
Przeczytaj pełny opis metody
Zaloguj się na bezpłatne konto, aby przeczytać tę sekcję.
Method map
The neighbourhood of related methods — select a node to explore.
Źródła
- Radford, A., Kim, J. W., Hallacy, C., Ramesh, A., Goh, G., Agarwal, S., ... & Sutskever, I. (2021). Learning transferable visual models from natural language supervision. In Proceedings of the 38th International Conference on Machine Learning (ICML), pp. 8748–8763. PMLR. link ↗
- Frome, A., Corrado, G. S., Shlens, J., Bengio, S., Dean, J., Ranzato, M., & Mikolov, T. (2013). DeViSE: A deep visual-semantic embedding model. In Advances in Neural Information Processing Systems (NeurIPS), Vol. 26. link ↗
Jak cytować tę stronę
ScholarGate. (2026, June 3). Multimodal Sentence Embeddings (Joint Vision-Language Representation Learning). ScholarGate. https://scholargate.app/pl/deep-learning/multimodal-sentence-embeddings
Which method?
Set this method beside its closest kin and read them side by side — the library lays the books on the table; the choice is yours.
Compare side by side →Cytowana przez
Widzisz błąd na tej stronie? Zgłoś go lub zaproponuj poprawkę →