AI Paper Atlas

Research map / Vision-language and multimodal / Robot action and navigation

Counterfactual Vision-and-Language Navigation via Adversarial Path Sampler

Tsu-Jui Fu, Xin Eric Wang, Matthew F. Peterson, Scott T. Grafton, Miguel P. Eckstein, William Yang Wang

ECCV 2020 · Vision-language and multimodal › Robot action and navigation · approach: Reasoning-enhanced vision-language navigation · 82 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add APS-sampled instead of randomly-sampled augmented pathsSeq2Seq · R2R val-seenSuccess Rate (SR): 43.7 -> 48.2Improved
add APS-sampled instead of randomly-sampled augmented pathsSpeaker-Follower · R2R val-seenSuccess Rate (SR): 66.4 -> 68.2Improved
add APS-sampled instead of randomly-sampled augmented pathsRCM · R2R val-seenSuccess Rate (SR): 61.9 -> 63.2Improved
add APS-sampled instead of randomly-sampled augmented pathsSeq2Seq · R2R val-unseenSuccess Rate (SR): 22.6 -> 24.2Improved
add APS-sampled instead of randomly-sampled augmented pathsSpeaker-Follower · R2R val-unseenSuccess Rate (SR): 36.1 -> 38.8Improved
add APS-sampled instead of randomly-sampled augmented pathsRCM · R2R val-unseenSuccess Rate (SR): 45.6 -> 47.7Improved
add APS-sampled instead of randomly-sampled augmented pathsSeq2Seq · R2R testSuccess Rate (SR): 21.0 -> 22.5Improved
add APS-sampled instead of randomly-sampled augmented pathsSpeaker-Follower · R2R testSuccess Rate (SR): 34.8 -> 36.1Improved
add APS-sampled instead of randomly-sampled augmented pathsRCM · R2R testSuccess Rate (SR): 44.5 -> 45.1Improved
add environment-based pre-exploration with APSSeq2Seq + augaps · R2R val-unseenSuccess Rate (SR): 24.2 -> 27.0Improved
add environment-based pre-exploration with APSSpeaker-Follower + augaps · R2R val-unseenSuccess Rate (SR): 38.8 -> 42.0Improved
add environment-based pre-exploration with APSRCM + augaps · R2R val-unseenSuccess Rate (SR): 47.7 -> 48.0Improved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy