Uporedite metode

	Mehanizam pažnje	LSTM	Multimodalni Transformer
Oblast	Duboko učenje	Duboko učenje	Duboko učenje
Porodica	Machine learning	Machine learning	Machine learning
Godina nastanka	2015	1997	2019–2021
Tvorac	Bahdanau, D.; Luong, M.T.	Hochreiter, S. & Schmidhuber, J.	Lu et al. (ViLBERT); Radford et al. (CLIP)
Tip	Neural attention layer (encoder-decoder)	Recurrent neural network (gated memory cell)	Cross-modal attention-based deep learning model
Temeljni izvor	Bahdanau, D., Cho, K. & Bengio, Y. (2015). Neural Machine Translation by Jointly Learning to Align and Translate. ICLR.	Hochreiter, S. & Schmidhuber, J. (1997). Long Short-Term Memory. Neural Computation, 9(8), 1735–1780.	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.
Drugi nazivi	Dikkat Mekanizması (Bahdanau / Luong Attention), dikkat mekanizmasi, neural attention, additive attention	LSTM (Uzun Kısa Dönem Bellek Ağı), long short-term memory, LSTM network, recurrent neural network with memory cells	multimodal attention model, cross-modal transformer, vision-language transformer, multi-modal fusion transformer
Srodne	5	5	5
Sažetak	The attention mechanism, introduced by Bahdanau, Cho and Bengio in 2015 and refined by Luong, Pham and Manning the same year, lets a sequence decoder dynamically learn which of the encoder's outputs to focus on at each step. Before the Transformer, it substantially improved machine-translation quality by freeing models from compressing an entire input into a single fixed vector.	LSTM (Long Short-Term Memory) is a recurrent neural network architecture, introduced by Sepp Hochreiter and Jürgen Schmidhuber in 1997, that can learn long-term dependencies in sequential data and is widely used for time-series and sequence prediction. It keeps an internal memory that lets information persist across many time steps.	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.