AI Paper Atlas

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

Conditional Image-Text Embedding Networks

Bryan A. Plummer, Paige Kordas, M. Hadi Kiapour, Shuai Zheng, Robinson Piramuthu, Svetlana Lazebnik

ECCV 2018 · Vision-language and multimodal › Human and embodied modeling · approach: Language-grounded scene text and structure · 123 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 spatial location featuresSimilarity Network · Flickr30K Entities test (PASCAL-tuned features)accuracy: 53.45 -> 54.52Improved
add learned concept weight branch instead of fixed K-means assignment (ablation)CITE · Flickr30K Entities test (PASCAL-tuned features)accuracy: 57.89 -> 58.69Improved
use K-means text-cluster assignments instead of random assignment (ablation)CITE · Flickr30K Entities test (PASCAL-tuned features)accuracy: 57.58 -> 57.89Improved
use random embedding assignment instead of coarse categories (ablation)CITE · Flickr30K Entities test (PASCAL-tuned features)accuracy: 55.42 -> 57.58Improved
share weights across conditional embeddings (CITE) instead of separate networks per coarse category (ablation)CITE · Flickr30K Entities test (PASCAL-tuned features)accuracy: 55.32 -> 55.42No clear effect
increase Edge Box proposals from 200 to 500CITE, Learned, K=4 · Flickr30K Entities test (PASCAL-tuned features)accuracy: 58.69 -> 59.27Improved
increase Edge Box proposals from 500 to 1000CITE, Learned, K=4 · Flickr30K Entities test (PASCAL-tuned features)accuracy: 59.27 -> 58.63Hurt
reduce embedding size from 256 to 64CITE, Learned, K=4 · Flickr30K Entities test (PASCAL-tuned features)accuracy: 58.69 -> 57.77Hurt
use Flickr30K-tuned features instead of PASCAL-tuned featuresCITE, Learned, K=4, 500 Edge Boxes · Flickr30K Entities testaccuracy: 59.27 -> 61.89Improved
use learned concept weights instead of K-means (ablation)CITE, K=12 · Visual Genome testaccuracy: 23.67 -> 24.43Improved
use learned concept weights instead of K-means, K=2 vs K=12 on ReferIt (ablation)CITE · ReferIt Game testaccuracy: 34.01 -> 34.13Mixed
use CITE K-means instead of Similarity Network baselineCITE, K-Means, K=12 · Visual Genome testaccuracy: 20.08 -> 23.67Improved

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