AI Paper Atlas

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

Language Features Matter: Effective Language Representations for Vision-Language Tasks

Andrea Burns, Reuben Tan, Kate Saenko, Stan Sclaroff, Bryan A. Plummer

ICCV 2019 · Vision-language and multimodal › Human and embodied modeling · approach: Vision-language pretraining · 30 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 pretrained Word2Vec instead of random scratch embedding (ablation)LSTM language model · Flickr30K image-sentence retrievalmean recall: 60.0 -> 68.5Improved
fine-tune Word2Vec Self-Attention embedding during training (ablation)Self-Attention language model · Flickr30K image-sentence retrievalmean recall: 63.6 -> 71.9Improved
use Self-Attention instead of LSTM on frozen Word2Vec (ablation)Embedding Network · Flickr30K image-sentence retrievalmean recall: 68.5 -> 63.6No clear effect
replace LSTM with Self-Attention language model (fine-tuned Word2Vec) (ablation)Embedding Network · Flickr30K image-sentence retrievalmean recall: 69.0 -> 71.9Improved
use BERT sentence embedding instead of fine-tuned Word2Vec (ablation)Embedding Network · Flickr30K image-sentence retrievalmean recall: 71.9 -> 71.8Hurt
use InferSent instead of BERT for Flickr30K Entities (ablation)CITE · Flickr30K Entities phrase groundingaccuracy: 69.38 -> 57.83Hurt
add WordNet and Visual Genome graph retrofitting (GrOVLE) to Word2VecSelf-Attention + ft · Flickr30K image-sentence retrievalmean recall: 71.9 -> 72.1Improved
use Visual Word2Vec instead of GrOVLE (ablation)Self-Attention + ft · Flickr30K image-sentence retrievalmean recall: 72.1 -> 68.8Hurt
add adapted embeddings (GrOVLE) to generation tasksARNet captioning · MSCOCO captioningCIDEr: 94.0 -> 92.5No clear effect
add multi-task pretraining w/ target task + fine-tuning to GrOVLE (ablation)Embedding Network · Flickr30K image-sentence retrievalmean recall: 64.7 -> 72.6Improved
add multi-task pretraining w/ target task + fine-tuning to GrOVLE (ablation)Embedding Network · MSCOCO image-sentence retrievalmean recall: 75.0 -> 81.3Improved
add multi-task pretraining w/ target task + fine-tuning to GrOVLE (ablation)CITE · ReferIt phrase groundingaccuracy: 52.15 -> 54.51Improved

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