Abstract
In this paper, we propose an error causal inference method that could be used for finding dominant features for a faulty instance under a well-trained multi-modality input model, which could apply to any testing instance. We evaluate our method using a well-trained multimodalities stylish caption generation model and find those causal inferences that could provide us the insights for next step optimization.
| Original language | English |
|---|---|
| Title of host publication | 2nd Workshop on Advances in Language and Vision Research, ALVR 2021 - Proceedings |
| Editors | Xin Wang, Ronghang Hu, Drew Hudson, Tsu-Jui Fu, Marcus Rohrbach, Daniel Fried |
| Pages | 11-15 |
| Number of pages | 5 |
| ISBN (Electronic) | 9781954085374 |
| State | Published - 2021 |
| Event | 2nd Workshop on Advances in Language and Vision Research, ALVR 2021 - Atlanta, United States Duration: Jun 11 2021 → … |
Publication series
| Name | 2nd Workshop on Advances in Language and Vision Research, ALVR 2021 - Proceedings |
|---|
Conference
| Conference | 2nd Workshop on Advances in Language and Vision Research, ALVR 2021 |
|---|---|
| Country/Territory | United States |
| City | Atlanta |
| Period | 6/11/21 → … |
Bibliographical note
Publisher Copyright:©2021 Association for Computational Linguistics
ASJC Scopus subject areas
- Computer Science Applications
- Software
- Linguistics and Language
- Language and Linguistics
- Ophthalmology