AI Paper Atlas

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

Weakly-Supervised Multi-Granularity Map Learning for Vision-and-Language Navigation

Peihao Chen, Dongyu Ji, Kunyang Lin, Runhao Zeng, Thomas H. Li, Mingkui Tan, Chuang Gan

NeurIPS 2022 · Vision-language and multimodal › Robot action and navigation · approach: Pretraining and data augmentation for VLN · 16 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
Use WS-MGMap instead of CM2VLN-CE navigation · VLN-CE Val-SeenSR: 42.9 -> 46.9Improved
Use WS-MGMap instead of CM2VLN-CE navigation · VLN-CE Val-UnseenSR: 34.3 -> 38.9Improved
Transfer the R2R-Habitat model without fine-tuningWS-MGMap · RxR-Habitat English Val-SeenSR: 9.0 -> 14.0Improved
Transfer the R2R-Habitat model without fine-tuningWS-MGMap · RxR-Habitat English Val-UnseenSR: 10.0 -> 15.0Improved
Add a multi-granularity map combining fine-grained and semantic information (ablation)WS-MGMap · VLN-CE Val-UnseenSR: 31.6 and 33.3 -> 38.9Improved
Add an explicit map instead of using raw RGB-D alone (ablation)WS-MGMap · VLN-CE Val-UnseenSR: 26.5 -> 38.9Improved
Add weakly supervised instruction-relevant object localization loss (ablation)WS-MGMap · VLN-CE Val-UnseenSR: 33.1 -> 38.9Improved
Use soft rather than hard localization ground truth (ablation)WS-MGMap localization task · VLN-CE Val-UnseenSR: 35.2 -> 38.9Improved
Use last-layer rather than first-layer U-Net features for the fine-grained map (ablation)WS-MGMap · VLN-CE Val-UnseenSR: 32.1 -> 38.9Improved
Use last-layer rather than classification-layer U-Net features for the fine-grained map (ablation)WS-MGMap · VLN-CE Val-UnseenSR: 33.3 -> 38.9Improved
Add semantic hallucination beyond the current field of view (ablation)WS-MGMap · VLN-CE Val-UnseenSR: 37.3 -> 38.9Improved
Use DAgger rather than imitation learning alone (ablation)WS-MGMap training · VLN-CE Val-UnseenSR: 24.6 -> 38.9Improved

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