AI Paper Atlas

Research map / Vision-language and multimodal / Visual question answering

Embodied Question Answering

Abhishek Das, Samyak Datta, Georgia Gkioxari, Stefan Lee, Devi Parikh, Dhruv Batra

CVPR 2018 · Vision-language and multimodal › Visual question answering · approach: Multimodal question answering datasets · 352 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 an ACT planner-controller instead of a standard LSTM navigator (ablation)EmbodiedQA navigation · EQA v1 test, T_-10d_T (m): 1.13 -> 0.46Improved
use an ACT planner-controller instead of a standard LSTM navigator (ablation)EmbodiedQA navigation · EQA v1 test, T_-30d_T (m): 2.23 -> 1.50Improved
use an ACT planner-controller instead of a standard LSTM navigator (ablation)EmbodiedQA navigation · EQA v1 test, T_-50d_T (m): 2.89 -> 2.74Improved
add question encoding to a reactive navigator (ablation)Reactive navigation · EQA v1 test, T_-10d_T (m): 2.09 -> 1.58Improved
add question encoding to an LSTM navigator (ablation)LSTM navigation · EQA v1 test, T_-10d_T (m): 1.75 -> 1.13Improved
add memory to a reactive navigator (ablation)Question-agnostic navigation · EQA v1 test, T_-10d_T (m): 2.09 -> 1.75Improved
add memory to a question-conditioned reactive navigator (ablation)Question-conditioned navigation · EQA v1 test, T_-10d_T (m): 1.58 -> 1.13Improved
fine-tune navigation with reinforcement learning and target-distance shaping (ablation)ACT+Q · EQA v1 test, T_-30mean rank: 3.13 -> 2.99Improved
fine-tune navigation with reinforcement learning and target-distance shaping (ablation)ACT+Q · EQA v1 test, T_-10mean rank: 3.09 -> 3.13Hurt
fine-tune navigation with reinforcement learning and target-distance shaping (ablation)ACT+Q · EQA v1 test, T_-50mean rank: 3.25 -> 3.22Improved
pretrain perception heads jointly rather than independently (ablation)CNN segmentation · Perception validation setmean IOU: 0.163 -> 0.166Improved
use human navigation views rather than shortest-path viewsQuestion answering module · Human navigationmean rank: 3.26 -> 3.51Hurt

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