IGLUE: A Benchmark for Transfer Learning across Modalities, Tasks, and Languages

Emanuele Bugliarello, Fangyu Liu, Jonas Pfeiffer, Siva Reddy, Desmond Elliott, Edoardo Maria Ponti, Ivan Vulić
Proceedings of the 39th International Conference on Machine Learning, PMLR 162:2370-2392, 2022.

Abstract

Reliable evaluation benchmarks designed for replicability and comprehensiveness have driven progress in machine learning. Due to the lack of a multilingual benchmark, however, vision-and-language research has mostly focused on English language tasks. To fill this gap, we introduce the Image-Grounded Language Understanding Evaluation benchmark. IGLUE brings together{—}by both aggregating pre-existing datasets and creating new ones{—}visual question answering, cross-modal retrieval, grounded reasoning, and grounded entailment tasks across 20 diverse languages. Our benchmark enables the evaluation of multilingual multimodal models for transfer learning, not only in a zero-shot setting, but also in newly defined few-shot learning setups. Based on the evaluation of the available state-of-the-art models, we find that translate-test transfer is superior to zero-shot transfer and that few-shot learning is hard to harness for many tasks. Moreover, downstream performance is partially explained by the amount of available unlabelled textual data for pretraining, and only weakly by the typological distance of target{–}source languages. We hope to encourage future research efforts in this area by releasing the benchmark to the community.

Cite this Paper

BibTeX
@InProceedings{pmlr-v162-bugliarello22a, title = {{IGLUE}: A Benchmark for Transfer Learning across Modalities, Tasks, and Languages}, author = {Bugliarello, Emanuele and Liu, Fangyu and Pfeiffer, Jonas and Reddy, Siva and Elliott, Desmond and Ponti, Edoardo Maria and Vuli{\'c}, Ivan}, booktitle = {Proceedings of the 39th International Conference on Machine Learning}, pages = {2370--2392}, year = {2022}, editor = {Chaudhuri, Kamalika and Jegelka, Stefanie and Song, Le and Szepesvari, Csaba and Niu, Gang and Sabato, Sivan}, volume = {162}, series = {Proceedings of Machine Learning Research}, month = {17--23 Jul}, publisher = {PMLR}, pdf = {https://proceedings.mlr.press/v162/bugliarello22a/bugliarello22a.pdf}, url = {https://proceedings.mlr.press/v162/bugliarello22a.html}, abstract = {Reliable evaluation benchmarks designed for replicability and comprehensiveness have driven progress in machine learning. Due to the lack of a multilingual benchmark, however, vision-and-language research has mostly focused on English language tasks. To fill this gap, we introduce the Image-Grounded Language Understanding Evaluation benchmark. IGLUE brings together{—}by both aggregating pre-existing datasets and creating new ones{—}visual question answering, cross-modal retrieval, grounded reasoning, and grounded entailment tasks across 20 diverse languages. Our benchmark enables the evaluation of multilingual multimodal models for transfer learning, not only in a zero-shot setting, but also in newly defined few-shot learning setups. Based on the evaluation of the available state-of-the-art models, we find that translate-test transfer is superior to zero-shot transfer and that few-shot learning is hard to harness for many tasks. Moreover, downstream performance is partially explained by the amount of available unlabelled textual data for pretraining, and only weakly by the typological distance of target{–}source languages. We hope to encourage future research efforts in this area by releasing the benchmark to the community.} }
Endnote
%0 Conference Paper %T IGLUE: A Benchmark for Transfer Learning across Modalities, Tasks, and Languages %A Emanuele Bugliarello %A Fangyu Liu %A Jonas Pfeiffer %A Siva Reddy %A Desmond Elliott %A Edoardo Maria Ponti %A Ivan Vulić %B Proceedings of the 39th International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2022 %E Kamalika Chaudhuri %E Stefanie Jegelka %E Le Song %E Csaba Szepesvari %E Gang Niu %E Sivan Sabato %F pmlr-v162-bugliarello22a %I PMLR %P 2370--2392 %U https://proceedings.mlr.press/v162/bugliarello22a.html %V 162 %X Reliable evaluation benchmarks designed for replicability and comprehensiveness have driven progress in machine learning. Due to the lack of a multilingual benchmark, however, vision-and-language research has mostly focused on English language tasks. To fill this gap, we introduce the Image-Grounded Language Understanding Evaluation benchmark. IGLUE brings together{—}by both aggregating pre-existing datasets and creating new ones{—}visual question answering, cross-modal retrieval, grounded reasoning, and grounded entailment tasks across 20 diverse languages. Our benchmark enables the evaluation of multilingual multimodal models for transfer learning, not only in a zero-shot setting, but also in newly defined few-shot learning setups. Based on the evaluation of the available state-of-the-art models, we find that translate-test transfer is superior to zero-shot transfer and that few-shot learning is hard to harness for many tasks. Moreover, downstream performance is partially explained by the amount of available unlabelled textual data for pretraining, and only weakly by the typological distance of target{–}source languages. We hope to encourage future research efforts in this area by releasing the benchmark to the community.
APA
Bugliarello, E., Liu, F., Pfeiffer, J., Reddy, S., Elliott, D., Ponti, E.M. & Vulić, I.. (2022). IGLUE: A Benchmark for Transfer Learning across Modalities, Tasks, and Languages. Proceedings of the 39th International Conference on Machine Learning, in Proceedings of Machine Learning Research 162:2370-2392 Available from https://proceedings.mlr.press/v162/bugliarello22a.html.

Related Material