Machine Translation @ UPC

SHAS

SHAS: Approaching optimal Segmentation for End-to-End Speech Translation

transformer-contributions

Measuring the Mixing of Contextual Information in the Transformer

ZeroSwot

Pushing the Limits of Zero-shot End-to-End Speech Translation

joint

Joint Source-Target Self Attention with Locality Constraints

transformer-contributions-nmt

iwslt-2021

Systems submitted to IWSLT 2021 by the MT-UPC group.

logit-explanations

iwslt-2022

Systems submitted to IWSLT 2022 by the MT-UPC group.

fairseq

Facebook AI Research Sequence-to-Sequence Toolkit written in Python.

speechalign

OccGen_dataset

OCCGEN dataset and its metadata

OccGen_toolkit

This is the project of creating balanced datasets (multilingual, bilingual and monolingual)

s2t-perceiver

Efficient Speech Translation with Dynamic Latent Perceivers

SegAugment

SEGAUGMENT: Maximizing the Utility of Speech Translation Data with Segmentation-based Augmentations

.github

blog

Blog of the MT-UPC group

ReSeTOX

Cod for the paper "ReSeTOX: Re-learning attention weights for toxicity mitigation in machine translation"

iwslt-2023

Systems submitted to IWSLT 2023 by the MT-UPC group.

publications

resum

PAI CCMA. Transcripció i resum. IDEAI-UPC