Towards Understanding Sample Variance in Visually Grounded Language Generation: Evaluations and Observations

Wanrong Zhu, Xin Wang, Pradyumna Narayana, Kazoo Sone, Sugato Basu, William Yang Wang


Abstract
A major challenge in visually grounded language generation is to build robust benchmark datasets and models that can generalize well in real-world settings. To do this, it is critical to ensure that our evaluation protocols are correct, and benchmarks are reliable. In this work, we set forth to design a set of experiments to understand an important but often ignored problem in visually grounded language generation: given that humans have different utilities and visual attention, how will the sample variance in multi-reference datasets affect the models’ performance? Empirically, we study several multi-reference datasets and corresponding vision-and-language tasks. We show that it is of paramount importance to report variance in experiments; that human-generated references could vary drastically in different datasets/tasks, revealing the nature of each task; that metric-wise, CIDEr has shown systematically larger variances than others. Our evaluations on reference-per-instance shed light on the design of reliable datasets in the future.
Anthology ID:
2020.emnlp-main.708
Volume:
Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP)
Month:
November
Year:
2020
Address:
Online
Venue:
EMNLP
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
8806–8811
Language:
URL:
https://aclanthology.org/2020.emnlp-main.708
DOI:
10.18653/v1/2020.emnlp-main.708
Bibkey:
Cite (ACL):
Wanrong Zhu, Xin Wang, Pradyumna Narayana, Kazoo Sone, Sugato Basu, and William Yang Wang. 2020. Towards Understanding Sample Variance in Visually Grounded Language Generation: Evaluations and Observations. In Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP), pages 8806–8811, Online. Association for Computational Linguistics.
Cite (Informal):
Towards Understanding Sample Variance in Visually Grounded Language Generation: Evaluations and Observations (Zhu et al., EMNLP 2020)
Copy Citation:
PDF:
https://preview.aclanthology.org/ingestion-script-update/2020.emnlp-main.708.pdf
Video:
 https://slideslive.com/38939350
Data
COCOFlickr30kVATEXVIST