Vision-and-Language Navigation via Causal Learning

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Wang, Liuyi, He, Zongtao, Dang, Ronghao, Shen, Mengjiao, Liu, Chengju, Chen, Qijun
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866913316536844288
author Wang, Liuyi
He, Zongtao
Dang, Ronghao
Shen, Mengjiao
Liu, Chengju
Chen, Qijun
author_facet Wang, Liuyi
He, Zongtao
Dang, Ronghao
Shen, Mengjiao
Liu, Chengju
Chen, Qijun
contents In the pursuit of robust and generalizable environment perception and language understanding, the ubiquitous challenge of dataset bias continues to plague vision-and-language navigation (VLN) agents, hindering their performance in unseen environments. This paper introduces the generalized cross-modal causal transformer (GOAT), a pioneering solution rooted in the paradigm of causal inference. By delving into both observable and unobservable confounders within vision, language, and history, we propose the back-door and front-door adjustment causal learning (BACL and FACL) modules to promote unbiased learning by comprehensively mitigating potential spurious correlations. Additionally, to capture global confounder features, we propose a cross-modal feature pooling (CFP) module supervised by contrastive learning, which is also shown to be effective in improving cross-modal representations during pre-training. Extensive experiments across multiple VLN datasets (R2R, REVERIE, RxR, and SOON) underscore the superiority of our proposed method over previous state-of-the-art approaches. Code is available at https://github.com/CrystalSixone/VLN-GOAT.
format Preprint
id arxiv_https___arxiv_org_abs_2404_10241
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Vision-and-Language Navigation via Causal Learning
Wang, Liuyi
He, Zongtao
Dang, Ronghao
Shen, Mengjiao
Liu, Chengju
Chen, Qijun
Computer Vision and Pattern Recognition
Artificial Intelligence
In the pursuit of robust and generalizable environment perception and language understanding, the ubiquitous challenge of dataset bias continues to plague vision-and-language navigation (VLN) agents, hindering their performance in unseen environments. This paper introduces the generalized cross-modal causal transformer (GOAT), a pioneering solution rooted in the paradigm of causal inference. By delving into both observable and unobservable confounders within vision, language, and history, we propose the back-door and front-door adjustment causal learning (BACL and FACL) modules to promote unbiased learning by comprehensively mitigating potential spurious correlations. Additionally, to capture global confounder features, we propose a cross-modal feature pooling (CFP) module supervised by contrastive learning, which is also shown to be effective in improving cross-modal representations during pre-training. Extensive experiments across multiple VLN datasets (R2R, REVERIE, RxR, and SOON) underscore the superiority of our proposed method over previous state-of-the-art approaches. Code is available at https://github.com/CrystalSixone/VLN-GOAT.
title Vision-and-Language Navigation via Causal Learning
topic Computer Vision and Pattern Recognition
Artificial Intelligence
url https://arxiv.org/abs/2404.10241