AI Paper Atlas

Research map / Vision-language and multimodal / Human and embodied modeling

Aligning Where to See and What to Tell: Image Captioning with Region-Based Attention and Scene-Specific Contexts

Kun Fu, Junqi Jin, Runpeng Cui, Fei Sha, Changshui Zhang

TPAMI 2016 · Vision-language and multimodal › Human and embodied modeling · approach: Concept-guided image captioning · 275 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 region-based attention (ablation)our-base-greedy · MSCOCOCIDEr-D: 70.7 -> 75.1Improved
add scene-factorized LSTMs (ablation)our-base-greedy · MSCOCOCIDEr-D: 70.7 -> 74.3Improved
add both region attention and scene factorization (ablation)our-base-greedy · MSCOCOCIDEr-D: 70.7 -> 78.3Improved
add scene factorization to region attention (ablation)our-ra-greedy · MSCOCOCIDEr-D: 75.1 -> 78.3Improved
add region attention to scene factorization (ablation)our-sf-greedy · MSCOCOCIDEr-D: 74.3 -> 78.3Improved
use beam search (beam 10) instead of greedyour-(ra+sf) · MSCOCOCIDEr-D: 78.3 -> 83.8Improved
use beam search (beam 10) instead of greedyour-(ra+sf) · MSCOCOBLEU-4: 24.6 -> 28.2Improved
add region attention to base (ablation)our-base-greedy · MSCOCOMETEOR: 20.0 -> 22.2Improved
add region attention instead of scene factorization (BLEU-4) (ablation)our-ra-greedy · MSCOCOBLEU-4: 24.2 -> 23.5Hurt
add region attention to base (ablation)our-base-greedy · MSCOCOBLEU-4: 22.6 -> 23.5Improved
distort scene vector to another topicour model · MSCOCO qualitativecaption contentImproved
use our-ra+sf-beam over mRNN for caption to image retrievalour-ra+sf-beam · MSCOCO retrievalR@1: 29.0 -> 29.3Improved

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