AI Paper Atlas

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

Learning to Set Waypoints for Audio-Visual Navigation

Changan Chen, Sagnik Majumder, Ziad Al-Halah, Ruohan Gao, Santhosh Kumar Ramakrishnan, Kristen Grauman

ICLR 2021 · Vision-language and multimodal › Robot action and navigation · approach: Pretraining and data augmentation for VLN · 36 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
learn audio-visual waypoints with end-to-end reinforcement learning and structured acoustic memoryAudioGoal navigation · Replica heardSPL: 78.2 -> 86.6Improved
learn audio-visual waypoints with end-to-end reinforcement learning and structured acoustic memoryAudioGoal navigation · Matterport3D heardSPL: 55.1 -> 72.3Improved
learn audio-visual waypoints with end-to-end reinforcement learning and structured acoustic memoryAudioGoal navigation · Replica unheardSNA: 16.7 -> 27.1Improved
learn audio-visual waypoints with end-to-end reinforcement learning and structured acoustic memoryAudioGoal navigation · Matterport3D unheardSPL: 25.9 -> 40.9Improved
add acoustic memory (ablation)AV-WaN · Matterport3D unheardSPL: 36.7 -> 40.9Improved
add acoustic memory (ablation)AV-WaN · Replica unheardSR: 63.3 -> 52.8Hurt
add geometric map (ablation)AV-WaN · Replica unheardSPL: 27.0 -> 34.7Improved
add geometric map (ablation)AV-WaN · Matterport3D unheardSPL: 25.4 -> 40.9Improved
add acoustic and geometric maps (ablation)AV-WaN · Matterport3D unheardSPL: 20.5 -> 40.9Improved
use learned waypoints instead of primitive actions (ablation)AV-WaN · Matterport3D heardSPL: 44.3 -> 72.3Improved
use learned waypoints instead of primitive actions (ablation)AV-WaN · Replica heardSPL: 79.8 -> 86.6Improved
add acoustic memory (ablation)AV-WaN · Replica heardSPL: 85.7 -> 86.6Improved

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