Machine learningDeep learning / NLP / CV

Multimodális transzformer

A multimodális transzformer (Multimodal Transformer) a standard transzformer architektúrát kiterjeszti két vagy több bemeneti modalitás – leggyakrabban szöveg és kép, de akár hang, videó vagy strukturált adat is – feldolgozására és közös következtetésre. A kereszt-modális figyelem (cross-modal attention) rétegek lehetővé teszik, hogy az egyik modalitásból származó információ befolyásolja a másikat, ami olyan feladatokat tesz lehetővé, mint a vizuális kérdés-válaszadás (visual question answering), képaláírás-generálás (image captioning) és multimodális szenzimentanalízis.

Megnyitás itt: MethodMindHamarosanVideóHamarosanDownload slides

A teljes módszer elolvasása

Csak tagoknak

Jelentkezzen be ingyenes fiókkal a szakasz elolvasásához.

Bejelentkezés

Method map

The neighbourhood of related methods — select a node to explore.

+15 more

Források

  1. 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. link
  2. 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. link

Hogyan hivatkozzon erre az oldalra

ScholarGate. (2026, June 3). Multimodal Transformer (Cross-Modal Attention-Based Architecture). ScholarGate. https://scholargate.app/hu/deep-learning/multimodal-transformer

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

Hivatkozik rá

ScholarGateMultimodal Transformer (Multimodal Transformer (Cross-Modal Attention-Based Architecture)). Letöltve 2026-06-15, forrás: https://scholargate.app/hu/deep-learning/multimodal-transformer · Adatkészlet: https://doi.org/10.5281/zenodo.20539026