Σύγκριση μεθόδων

	Πολυτροπικό Word2Vec	Πολυτροπικός Μετασχηματιστής
Πεδίο	Βαθιά Μάθηση	Βαθιά Μάθηση
Οικογένεια	Machine learning	Machine learning
Έτος προέλευσης	2014	2019–2021
Δημιουργός	Bruni, E., Tran, N.-K., & Baroni, M. (building on Mikolov et al.)	Lu et al. (ViLBERT); Radford et al. (CLIP)
Τύπος	Multimodal word embedding model	Cross-modal attention-based deep learning model
Θεμελιώδης πηγή	Bruni, E., Tran, N.-K., & Baroni, M. (2014). Multimodal Distributional Semantics. Journal of Artificial Intelligence Research, 49, 1–47.	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.
Εναλλακτικές ονομασίες	multimodal word embeddings, visual-linguistic Word2Vec, cross-modal Word2Vec, MM-W2V	multimodal attention model, cross-modal transformer, vision-language transformer, multi-modal fusion transformer
Συναφείς	5	5
Σύνοψη	Multimodal Word2Vec extends the classic Word2Vec framework by grounding word representations in perceptual signals — typically image features — alongside distributional text statistics. The result is word vectors that capture both linguistic co-occurrence patterns and visual meaning, enabling richer semantic similarity judgements and better performance on concept-level tasks where purely text-based embeddings fall short.	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.