AI Paper Atlas

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

Aligned Image-Word Representations Improve Inductive Transfer Across Vision-Language Tasks

Tanmay Gupta, Kevin Shih, Saurabh Singh, Derek Hoiem

ICCV 2017 · Vision-language and multimodal › Visual question answering · approach: Vision-language model VQA evaluation · 20 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 Genome classifiers jointly trained with VQA (ablation)VQA model · VQA valoverall accuracy: 62.9 -> 64.1Improved
use shared word representations for visual classification instead of a standard classification layer (ablation)joint VQA + Genome model · VQA valoverall accuracy: 64.1 -> 65.3Improved
add shared vision-language representation with Genome supervision (ablation)VQA model · VQA val, what coloraccuracy: 53.5 -> 62.1Improved
add shared representation with Genome supervision (ablation)VQA model · VQA val, numberaccuracy: 39.2 -> 38.8No clear effect
add shared representation with Genome supervision (ablation)VQA model · VQA val, why/howaccuracy: 33.8 -> 33.6Mixed
add relevance features (ablation)VQA + Genome (Shared) · VQA valoverall accuracy: 65.3 -> 65.5Improved
add binary features for answers (ablation)VQA + Genome (Shared) with relevance features · VQA valoverall accuracy: 65.5 -> 65.7Improved
pretrain classifiers on Genome before joint training instead of training jointly from scratch (ablation)VQA + Genome (Shared) · VQA valoverall accuracy: 65.3 -> 63.7Hurt
train with high weight on recognition losses and VQA as weak supervisionjoint model · Visual Genome testobject classification accuracyImproved

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