AI Paper Atlas

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

Interpretable Visual Question Answering by Reasoning on Dependency Trees

Qingxing Cao, Xiaodan Liang, Bailin Li, Liang Lin

TPAMI 2019 · Vision-language and multimodal › Visual question answering · approach: Attention and prior-robust VQA · 58 citations (OpenAlex)

Read the paper

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 edge-dependent propagation (ablation)baseline PTGRN · CLEVR valoverall accuracy: 87.62 -> 89.28Improved
add GRU gated residual composition to +Edge (ablation)PTGRN +Edge · CLEVR valoverall accuracy: 89.28 -> 90.44Improved
add GRU gating when attention map is concatenated into hidden (ablation)PTGRN +Edge+ConcatAtt · CLEVR valoverall accuracy: 93.36 -> 93.53No clear effect
process attention map with a separate GRU instead of concatenating with hidden (ablation)PTGRN +Edge+GRU · CLEVR valoverall accuracy: 93.53 -> 94.42Improved
use convolutional GRU for attention map encoding to preserve spatial information (ablation)PTGRN +Edge+GRU+GRUAtt · CLEVR valoverall accuracy: 94.42 -> 95.42Improved
use Bi-LSTM context word encoding instead of lookup table (ablation)PTGRN · CLEVR valoverall accuracy: 87.75 -> 95.42Improved
use ground-truth program layout instead of parsed dependency tree (ablation)PTGRN · CLEVR valoverall accuracy: 95.42 -> 96.69Improved
randomly perturb dependency relations (10% to 70%) (ablation)PTGRN · CLEVR valoverall accuracy: 95.42 -> 86.77 (10%), 40.11 (70%)Hurt
redeveloped modules (shared weights, edge-dependent, gated, ConvGRU attention) versus preliminary ACMNPTGRN · CLEVR testoverall accuracy: 89.3 -> 95.5Improved
PTGRN versus MAC on attention accuracy without ground-truth layoutsPTGRN · CLEVR valattention accuracy: 32.72 -> 64.71Improved
PTGRN versus FiLM on unseen color combinationsPTGRN · CLEVR-CoGenT val (train A, test B)accuracy: 75.6 -> 83.50Improved
PTGRN versus ACMN on natural imagesPTGRN · VQAv2 test-dev (All)accuracy: 63.81 -> 67.07Improved

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