Presentation | 2022-12-16 On a difficulty of TextVQA Koki Nakamura, Seiichi Uchida, |
---|---|
PDF Download Page | PDF download Page Link |
Abstract(in Japanese) | (See Japanese page) |
Abstract(in English) | Text Visual Question Answering (TextVQA) is a task that uses textual information in images to generate natural language answers to questions about the images. Various methods have been proposed for this task, but even with the current state-of-the-art, the accuracy rate is only about 70%. To understand what factors make TextVQA difficult, we attempted to analyze the results of binary classification predictions of whether or not the model is capable of producing correct answers, using the answers to the TextVQA model and the dataset used in its training as input. |
Keyword(in Japanese) | (See Japanese page) |
Keyword(in English) | TextVQA / multimodal / binary classification |
Paper # | PRMU2022-51 |
Date of Issue | 2022-12-08 (PRMU) |
Conference Information | |
Committee | PRMU |
---|---|
Conference Date | 2022/12/15(2days) |
Place (in Japanese) | (See Japanese page) |
Place (in English) | Toyama International Conference Center |
Topics (in Japanese) | (See Japanese page) |
Topics (in English) | |
Chair | Seiichi Uchida(Kyushu Univ.) |
Vice Chair | Takuya Funatomi(NAIST) / Mitsuru Anpai(Denso IT Lab.) |
Secretary | Takuya Funatomi(CyberAgent) / Mitsuru Anpai(Univ. of Tokyo) |
Assistant | Nakamasa Inoue(Tokyo Inst. of Tech.) / Yasutomo Kawanishi(Riken) |
Paper Information | |
Registration To | Technical Committee on Pattern Recognition and Media Understanding |
---|---|
Language | JPN |
Title (in Japanese) | (See Japanese page) |
Sub Title (in Japanese) | (See Japanese page) |
Title (in English) | On a difficulty of TextVQA |
Sub Title (in English) | |
Keyword(1) | TextVQA |
Keyword(2) | multimodal |
Keyword(3) | binary classification |
1st Author's Name | Koki Nakamura |
1st Author's Affiliation | Kyushu University(Kyushu Univ.) |
2nd Author's Name | Seiichi Uchida |
2nd Author's Affiliation | Kyushu University(Kyushu Univ.) |
Date | 2022-12-16 |
Paper # | PRMU2022-51 |
Volume (vol) | vol.122 |
Number (no) | PRMU-314 |
Page | pp.pp.100-105(PRMU), |
#Pages | 6 |
Date of Issue | 2022-12-08 (PRMU) |