Papers by Meera Hahn
Where Are You? Localization from Embodied Dialog (2020.emnlp-main)
Copied to clipboard
| Challenge: | Observer and Locator perform a cooperative localization task in a 3D environment. |
| Approach: | They propose a dataset of 6k dialogs in which two humans complete a cooperative localization task. |
| Outcome: | The proposed model achieves 32.7% success at identifying the Observer’s location within 3m in unseen buildings, vs. 70.4% for human Locators. |
Transformer-based Localization from Embodied Dialog with Large-scale Pre-training (2022.aacl-short)
Copied to clipboard
| Challenge: | Observer agent and Locator agents communicate via natural language dialog without access to the Observers view of the scene. |
| Approach: | They propose a graph-based scene representation that is more effective than top-down 2D maps used in previous works. |
| Outcome: | The proposed model outperforms top-down 2D maps and is more accurate than previous baselines. |