VLUE: A Multi-Task Multi-Dimension Benchmark for Evaluating Vision-Language Pre-training

Wangchunshu Zhou, Yan Zeng, Shizhe Diao, Xinsong Zhang
Proceedings of the 39th International Conference on Machine Learning, PMLR 162:27395-27411, 2022.

Abstract

Recent advances in vision-language pre-training (VLP) have demonstrated impressive performance in a range of vision-language (VL) tasks. However, there exist several challenges for measuring the community’s progress in building general multi-modal intelligence. First, most of the downstream VL datasets are annotated using raw images that are already seen during pre-training, which may result in an overestimation of current VLP models’ generalization ability. Second, recent VLP work mainly focuses on absolute performance but overlooks the efficiency-performance trade-off, which is also an important indicator for measuring progress. To this end, we introduce the Vision-Language Understanding Evaluation (VLUE) benchmark, a multi-task multi-dimension benchmark for evaluating the generalization capabilities and the efficiency-performance trade-off (“Pareto SOTA”) of VLP models. We demonstrate that there is a sizable generalization gap for all VLP models when testing on out-of-distribution test sets annotated on images from a more diverse distribution that spreads across cultures. Moreover, we find that measuring the efficiency-performance trade-off of VLP models leads to complementary insights for several design choices of VLP. We release the VLUE benchmark to promote research on building vision-language models that generalize well to images unseen during pre-training and are practical in terms of efficiency-performance trade-off.

Cite this Paper

BibTeX
@InProceedings{pmlr-v162-zhou22n, title = {{VLUE}: A Multi-Task Multi-Dimension Benchmark for Evaluating Vision-Language Pre-training}, author = {Zhou, Wangchunshu and Zeng, Yan and Diao, Shizhe and Zhang, Xinsong}, booktitle = {Proceedings of the 39th International Conference on Machine Learning}, pages = {27395--27411}, 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/zhou22n/zhou22n.pdf}, url = {https://proceedings.mlr.press/v162/zhou22n.html}, abstract = {Recent advances in vision-language pre-training (VLP) have demonstrated impressive performance in a range of vision-language (VL) tasks. However, there exist several challenges for measuring the community’s progress in building general multi-modal intelligence. First, most of the downstream VL datasets are annotated using raw images that are already seen during pre-training, which may result in an overestimation of current VLP models’ generalization ability. Second, recent VLP work mainly focuses on absolute performance but overlooks the efficiency-performance trade-off, which is also an important indicator for measuring progress. To this end, we introduce the Vision-Language Understanding Evaluation (VLUE) benchmark, a multi-task multi-dimension benchmark for evaluating the generalization capabilities and the efficiency-performance trade-off (“Pareto SOTA”) of VLP models. We demonstrate that there is a sizable generalization gap for all VLP models when testing on out-of-distribution test sets annotated on images from a more diverse distribution that spreads across cultures. Moreover, we find that measuring the efficiency-performance trade-off of VLP models leads to complementary insights for several design choices of VLP. We release the VLUE benchmark to promote research on building vision-language models that generalize well to images unseen during pre-training and are practical in terms of efficiency-performance trade-off.} }
Endnote
%0 Conference Paper %T VLUE: A Multi-Task Multi-Dimension Benchmark for Evaluating Vision-Language Pre-training %A Wangchunshu Zhou %A Yan Zeng %A Shizhe Diao %A Xinsong Zhang %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-zhou22n %I PMLR %P 27395--27411 %U https://proceedings.mlr.press/v162/zhou22n.html %V 162 %X Recent advances in vision-language pre-training (VLP) have demonstrated impressive performance in a range of vision-language (VL) tasks. However, there exist several challenges for measuring the community’s progress in building general multi-modal intelligence. First, most of the downstream VL datasets are annotated using raw images that are already seen during pre-training, which may result in an overestimation of current VLP models’ generalization ability. Second, recent VLP work mainly focuses on absolute performance but overlooks the efficiency-performance trade-off, which is also an important indicator for measuring progress. To this end, we introduce the Vision-Language Understanding Evaluation (VLUE) benchmark, a multi-task multi-dimension benchmark for evaluating the generalization capabilities and the efficiency-performance trade-off (“Pareto SOTA”) of VLP models. We demonstrate that there is a sizable generalization gap for all VLP models when testing on out-of-distribution test sets annotated on images from a more diverse distribution that spreads across cultures. Moreover, we find that measuring the efficiency-performance trade-off of VLP models leads to complementary insights for several design choices of VLP. We release the VLUE benchmark to promote research on building vision-language models that generalize well to images unseen during pre-training and are practical in terms of efficiency-performance trade-off.
APA
Zhou, W., Zeng, Y., Diao, S. & Zhang, X.. (2022). VLUE: A Multi-Task Multi-Dimension Benchmark for Evaluating Vision-Language Pre-training. Proceedings of the 39th International Conference on Machine Learning, in Proceedings of Machine Learning Research 162:27395-27411 Available from https://proceedings.mlr.press/v162/zhou22n.html.

Related Material