2021 Fiscal Year Research-status Report
Linking Vision and Language through Computational Modelling
| Project/Area Number |
19K12733
|
|---|
| Research Institution | Kobe City University of Foreign Studies |
|---|
Principal Investigator |
CHANG Franklin 神戸市外国語大学, 外国語学部, 教授 (60827343)
|
|---|
| Project Period (FY) |
2019-04-01 – 2024-03-31
|
| Keywords | Vision / Language / Learning / Syntax / Computational model / Deep Learning / Priming |
|---|
| Outline of Annual Research Achievements |
I am investigating prediction error based learning, which is an important feature of deep learning algorithms that I am using for the modelling work. To examine this, several structural priming studies have been done to see whether learners are sensitive to prediction error.
I have developed a deep learning model which learns to map between visual inputs and Japanese verbs has been developed. Experiments have been done to test the predictions of the model.
|
| Current Status of Research Progress |
Current Status of Research Progress
3: Progress in research has been slightly delayed.
Reason
Due to changes in the research plan and the corona pandemic, the research program has shifted towards online experiments. This has required some change in the research plans.
Now work on the deep learning model has begun, so hopefully it will progress more smoothly from here.
|
| Strategy for Future Research Activity |
I plan to publish a research paper on the deep learning verb production model within a few months. Then I plan to work on a more comprehensive model that can deal with thematic roles and other aspects of production.
|
| Causes of Carryover |
I was not able to use up all of the money and so I would like to use it next year. I plan to do more computational modeling and then test the model with human experiments.
|
