AI Paper Atlas

Research map / Video / Video-language understanding

Env-QA: A Video Question Answering Benchmark for Comprehensive Understanding of Dynamic Environments

Difei Gao, Ruiping Wang, Ziyi Bai, Xilin Chen

ICCV 2021 · Video › Video-language understanding · approach: Video question answering · 26 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 multi-step attention (ablation)TSEA · Env-QA test splitoverall accuracy (%): 45.46 -> 47.06Improved
add event-level feature (temporal segmentation) (ablation)TSEA without multi-step attention · Env-QA test splitoverall accuracy (%): 43.39 -> 45.46Improved
add focus attention (ablation)TSEA without event feature · Env-QA test splitoverall accuracy (%): 41.80 -> 43.39Improved
add object name feature (ablation)TSEA without focus attention · Env-QA test splitoverall accuracy (%): 40.60 -> 41.80Improved
evaluate on unseen environments instead of seen environments (ablation)TSEA · Env-QA test splitaccuracy (%): 46.85 -> 47.20No clear effect
use TSEA instead of STAGEvideo QA model · Env-QA test splitoverall accuracy (%): 42.53 -> 47.06Improved
add visual input to question (CNN-LSTM versus Q-ONLY)video QA baseline · Env-QA test splitoverall accuracy (%): 32.48 -> 38.05Improved
use TSEA instead of Q-ONLY on Event questionsvideo QA · Env-QA test split, Event questionsaccuracy (%): 24.26 -> 39.84Improved
use TSEA instead of Q-ONLY on Number questionsvideo QA · Env-QA test split, Number questionsaccuracy (%): 37.84 -> 39.35No clear effect
use TSEA instead of Q-ONLY on Order questionsvideo QA · Env-QA test split, Order questionsaccuracy (%): 51.79 -> 55.53No clear effect
use TSEA instead of STAGE on State questionsvideo QA · Env-QA test split, State questionsaccuracy (%): 49.93 -> 56.73Improved
use TSEA instead of STAGE on Action rolevideo QA · Env-QA test split, Action rolerole accuracy (%): 45.69 -> 47.61No clear effect

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