AI Paper Atlas

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

Embodied Navigation with Auxiliary Task of Action Description Prediction

Haru Kondoh, Asako Kanezaki

ICCV 2025 · Vision-language and multimodal › Robot action and navigation · approach: Pretraining and data augmentation for VLN · 1 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 Past-DescRL auxiliary taskSAVi · SAVNav heardSR: 31.6 -> 37.4Improved
add Past-DescRL auxiliary taskSAVi · SAVNav unheardSR: 24.7 -> 31.4Improved
add Future-DescRL auxiliary taskSAVi · SAVNav unheardSR: 24.7 -> 22.5Hurt
add Past-DescRL auxiliary taskKSAVEN · SAVNav heardSR: 25.1 -> 21.4Hurt
add Past-DescRL auxiliary taskSMT · ObjNavSR: 17.9 -> 26.7Improved
add Past-DescRL auxiliary taskDUET · VLN Val UnseenSPL: 60.44 -> 61.37Improved
add Past-DescRL auxiliary taskScaleVLN · VLN Val UnseenSPL: 69.15 -> 68.96No clear effect
add ADPredictor pre-training (ablation)SMT w/ Past-DescRL · ObjNavSR: 12.8 -> 26.7Improved
add ADPredictor pre-training (ablation)SAVi w/ DescRL · SAVNav heardSR: 33.9 -> 37.6Improved
add task embedding (ablation)SAVi w/ P-DescRL · SAVNav unheardSR: 30.0 -> 31.4Improved
share all 3 decoders instead of 2 (ablation)SAVi w/ P-DescRL · SAVNav unheardSR: 31.4 -> 26.5Hurt
use Past-DescRL instead of Next Action auxiliary taskSAVi · SAVNav unheardSR: 23.1 -> 31.4Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

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