AI Paper Atlas

Research map / Video / Video segmentation and tracking

(2.5+1)D Spatio-Temporal Scene Graphs for Video Question Answering

Anoop Cherian, Chiori Hori, Tim K. Marks, Jonathan Le Roux

AAAI 2022 · Video › Video segmentation and tracking · approach: Layered video decomposition and dynamics · 37 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 the (2.5+1)D-Transformer instead of graph-alignment VQAvideo QA · NExT-QA validationaccuracy (%): 49.74 -> 53.4Improved
use the (2.5+1)D-Transformer instead of ST Scene Graphsvideo QA · AVSD-QA testmean rank: 5.91 -> 5.84Improved
remove the dynamic graph (ablation)full (2.5+1)D-Transformer · NExT-QA validationaccuracy (%): 52.49 -> 53.4Hurt
remove the static graph (ablation)full (2.5+1)D-Transformer · NExT-QA validationaccuracy (%): 53.0 -> 53.4No clear effect
remove I3D motion features (ablation)full (2.5+1)D-Transformer · NExT-QA validationaccuracy (%): 52.65 -> 53.4Hurt
remove augmented-answer training (ablation)full (2.5+1)D-Transformer · NExT-QA validationaccuracy (%): 49.98 -> 53.4Hurt
remove question conditioning (ablation)full (2.5+1)D-Transformer · NExT-QA validationaccuracy (%): 50.39 -> 53.4Hurt
remove the hierarchical kernel (ablation)full (2.5+1)D-Transformer · NExT-QA validationaccuracy (%): 52.9 -> 53.4Hurt
add pseudo-depth to the no-depth V(2+1)D Transformer (ablation)NExT-QA video QA · NExT-QA validationaccuracy (%): 52.4 -> 53.4Improved
use all graph nodes instead of pruning (ablation)(2.5+1)D-Transformer · NExT-QA validationaccuracy (%): 53.4 -> 53.5Improved
increase hierarchy levels from 1 to 4 using the listed bandwidths (ablation)(2.5+1)D-Transformer · NExT-QA validationaccuracy: 52.13 -> 53.20Improved
increase hierarchy levels from 4 to 5 (ablation)(2.5+1)D-Transformer · NExT-QA validationaccuracy: 53.20 -> 53.0Hurt

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