Task Formulation Matters When Learning Continually: A Case Study in Visual Question Answering
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2022
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866917570830925824 |
|---|---|
| author | Nikandrou, Mavina Yu, Lu Suglia, Alessandro Konstas, Ioannis Rieser, Verena |
| author_facet | Nikandrou, Mavina Yu, Lu Suglia, Alessandro Konstas, Ioannis Rieser, Verena |
| contents | Continual learning aims to train a model incrementally on a sequence of tasks without forgetting previous knowledge. Although continual learning has been widely studied in computer vision, its application to Vision+Language tasks is not that straightforward, as settings can be parameterized in multiple ways according to their input modalities. In this paper, we present a detailed study of how different settings affect performance for Visual Question Answering. We first propose three plausible task formulations and demonstrate their impact on the performance of continual learning algorithms. We break down several factors of task similarity, showing that performance and sensitivity to task order highly depend on the shift of the output distribution. We also investigate the potential of pretrained models and compare the robustness of transformer models with different visual embeddings. Finally, we provide an analysis interpreting model representations and their impact on forgetting. Our results highlight the importance of stabilizing visual representations in deeper layers. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2210_00044 |
| institution | arXiv |
| publishDate | 2022 |
| record_format | arxiv |
| spellingShingle | Task Formulation Matters When Learning Continually: A Case Study in Visual Question Answering Nikandrou, Mavina Yu, Lu Suglia, Alessandro Konstas, Ioannis Rieser, Verena Machine Learning Continual learning aims to train a model incrementally on a sequence of tasks without forgetting previous knowledge. Although continual learning has been widely studied in computer vision, its application to Vision+Language tasks is not that straightforward, as settings can be parameterized in multiple ways according to their input modalities. In this paper, we present a detailed study of how different settings affect performance for Visual Question Answering. We first propose three plausible task formulations and demonstrate their impact on the performance of continual learning algorithms. We break down several factors of task similarity, showing that performance and sensitivity to task order highly depend on the shift of the output distribution. We also investigate the potential of pretrained models and compare the robustness of transformer models with different visual embeddings. Finally, we provide an analysis interpreting model representations and their impact on forgetting. Our results highlight the importance of stabilizing visual representations in deeper layers. |
| title | Task Formulation Matters When Learning Continually: A Case Study in Visual Question Answering |
| topic | Machine Learning |
| url | https://arxiv.org/abs/2210.00044 |