Machine learningDeep learning / NLP / CV

다중 모달 트랜스포머

다중 모달 트랜스포머(Multimodal Transformer)는 표준 트랜스포머 아키텍처를 확장하여 두 개 이상의 입력 모달리티(가장 일반적으로 텍스트와 이미지이지만 오디오, 비디오 또는 구조화된 데이터도 포함)를 처리하고 공동으로 추론합니다. 교차 모달 어텐션(cross-modal attention) 레이어는 한 모달리티의 정보가 다른 모달리티의 표현에 영향을 미치도록 하여 시각적 질문 응답, 이미지 캡셔닝, 다중 모달 감성 분석과 같은 작업을 가능하게 합니다.

MethodMind에서 열기곧 제공동영상곧 제공Download slides

방법 전문 읽기

회원 전용

무료 계정으로 로그인하면 이 섹션을 읽을 수 있습니다.

로그인

Method map

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

+15 more

출처

  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

이 페이지 인용 방법

ScholarGate. (2026, June 3). Multimodal Transformer (Cross-Modal Attention-Based Architecture). ScholarGate. https://scholargate.app/ko/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

이 방법을 참조하는 항목

ScholarGateMultimodal Transformer (Multimodal Transformer (Cross-Modal Attention-Based Architecture)). 2026-06-15에 다음에서 검색함: https://scholargate.app/ko/deep-learning/multimodal-transformer · 데이터셋: https://doi.org/10.5281/zenodo.20539026