Back to Search
Start Over
PASTS: Progress-aware spatio-temporal transformer speaker for vision-and-language navigation.
- Source :
-
Engineering Applications of Artificial Intelligence . Feb2024, Vol. 128, pN.PAG-N.PAG. 1p. - Publication Year :
- 2024
-
Abstract
- Vision-and-language navigation (VLN) is a crucial but challenging cross-modal navigation task. One powerful technique to enhance the generalization performance in VLN is the use of an independent speaker model to provide pseudo instructions for data augmentation. However, current speaker models based on Long-Short Term Memory (LSTM) lack the ability to attend to features relevant at different locations and time steps. To address this, we propose a novel progress-aware spatio-temporal transformer speaker (PASTS) model that uses the transformer as the core of the network. PASTS uses a spatio-temporal encoder to fuse panoramic representations and encode intermediate connections through steps. Besides, to avoid the misalignment problem that could result in incorrect supervision, a speaker progress monitor (SPM) is proposed to enable the model to estimate the progress of instruction generation and facilitate more fine-grained caption results. Additionally, a multifeature dropout (MFD) strategy is introduced to alleviate overfitting. The proposed PASTS is flexible to be combined with existing VLN models. The experimental results demonstrate that PASTS outperforms previous speaker models and successfully improves the performance of previous VLN models, achieving state-of-the-art performance on the standard Room-to-Room (R2R) dataset. • Spatio-temporal transformer encoder extracts trajectory features effectively. • Speaker progress monitor improves instruction alignment and generation. • Multifeature dropout strategy boosts generalization in unseen environments. • Stronger speaker model provides more accurate pseudo labels for data augmentation. [ABSTRACT FROM AUTHOR]
- Subjects :
- *TRANSFORMER models
*DATA augmentation
*PERFORMANCE standards
Subjects
Details
- Language :
- English
- ISSN :
- 09521976
- Volume :
- 128
- Database :
- Academic Search Index
- Journal :
- Engineering Applications of Artificial Intelligence
- Publication Type :
- Academic Journal
- Accession number :
- 174339425
- Full Text :
- https://doi.org/10.1016/j.engappai.2023.107487