Composed of an encoder (which compresses the input) and a decoder (which decompresses the compressed input).
Encoder-decoder architectures can handle inputs and outputs that both consist of variable-length sequences and thus are suitable for sequence-to-sequence problems such as machine translation. The encoder takes a variable-length sequence as input and transforms it into a state with a fixed shape. The decoder maps the encoded state of a fixed shape to a variable-length sequence.
Note
This architecture is not just used forĀ sequence transductionĀ tasks (i.e. language translation, summarization, image captioning etc). Example: variational autoencoder which is used in computer vision, but has an encoder-decoder architecture.