AI Paper Atlas

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

Deep Visual-Semantic Alignments for Generating Image Descriptions

Andrej Karpathy, Li Fei-Fei

TPAMI 2016 · Vision-language and multimodal › Human and embodied modeling · approach: Language-grounded scene text and structure · 4,990 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 simpler ranking cost instead of DeFrag loss (dependency tree edges) (ablation)DeFrag alignment · Flickr30K rankingImage annotation R@1: 19.2 -> 20.0Improved
replace dependency tree edges with BRNN (ablation)alignment model · Flickr30K rankingImage annotation R@1: 20.0 -> 22.2Improved
replace dependency tree edges with BRNN (ablation)alignment model · Flickr30K rankingImage search R@1: 15.0 -> 15.2Improved
replace dependency tree edges with BRNN (ablation)alignment model · Flickr8K rankingImage annotation R@1: 14.8 -> 16.5Improved
simpler cost function with dependency edges (ablation)DeFrag alignment · Flickr8K rankingImage annotation R@1: 13.8 -> 14.8Improved
use region-level model instead of full-frame modelmultimodal RNN · Region descriptionBLEU-1: 14.2 -> 35.2Improved
use multimodal RNN instead of nearest neighbor retrievalcaption generator · MSCOCO 2014 captioningCIDEr: 38.3 -> 66.0Improved
use multimodal RNN instead of nearest neighbor retrievalcaption generator · MSCOCO 2014 captioningBLEU-4: 10.0 -> 23.0Improved
use beam size 7 instead of beam size 1 (ablation)multimodal RNN · MSCOCO 2014 captioningCIDEr: 0.61 -> 0.66Improved
use LSTM instead of RNN (ablation)multimodal RNN · captioningqualityImproved
train word vectors from random init instead of word2vec (ablation)multimodal RNN · captioningqualityNo clear effect
clip gradients (ablation)multimodal RNN · captioningvalidation performanceImproved

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