Starter code for the VMT task and challenge

This code repository is for the accepted ACL2022 paper "On Vision Features in Multimodal Machine Translation". We provide the details and scripts for the proposed probing tasks. We hope the code could help those who want to research on the multimodal machine translation task.

Code and Data for the ACL22 main conference paper "MSCTD: A Multimodal Sentiment Chat Translation Dataset"

Code for the ACL 2023 paper Scene Graph as Pivoting: Inference-time Image-free Unsupervised Multimodal Machine Translation with Visual Scene Hallucination

ViTA: Visual-Linguistic Translation by Aligning Object Tags

Resources for my research on Multimodal Machine Translation within the framework of Frame Semantics (this is a WIP 🚧)