AI Paper Atlas

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

Generating Diverse and Descriptive Image Captions Using Visual Paraphrases

Lixin Liu, Jiajun Tang, Xiaojun Wan, Zongming Guo

ICCV 2019 · Vision-language and multimodal › Human and embodied modeling · approach: Neural image captioning · 43 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 the first decoding stage (two-step) instead of training only on second-stage captions (ablation)Ours (Tdiv, 0.1) · MS COCO testDist-3: 6957 -> 7810Improved
add the first decoding stage (two-step) instead of training only on second-stage captions (ablation)Ours (Tdiv, 0.1) · MS COCO testCIDEr: 102.6 -> 105.4Improved
add the first decoding stage over pretrain-then-finetune on second-stage captions (ablation)Ours (Tdiv, 0.1) · MS COCO testSPICE: 20.3 -> 21.0Improved
add the first decoding stage over pretrain-then-finetune on second-stage captions (ablation)Ours (Tdiv, 0.1) · MS COCO testR@1: 23.6 -> 25.0Improved
increase epsilon from 0.1 to 0.3 in TF-IDF diversity pair selectionOurs (Tdiv) · MS COCO testDist-3: 7810 -> 10053Improved
increase epsilon from 0.1 to 0.3 in TF-IDF diversity pair selectionOurs (Tdiv) · MS COCO testCIDEr: 105.4 -> 86.9Hurt
increase epsilon from 0.1 to 0.3 in TF-IDF diversity pair selectionOurs (Tdiv) · MS COCO human evaluationDiversity: 3.54 -> 4.06Improved
select pairs by sentence length instead of TdivOurs (len, 0) · MS COCO testCIDEr: 109.8 -> 69.7Hurt
select pairs by Yngve syntactic complexityOurs (Yngve, 0) · MS COCO testR@1: 19.2 -> 22.6Improved
select pairs by image retrieval rankOurs (IR, 2) · MS COCO testR@1: 19.2 -> 25.3Improved
select pairs by TF-IDF diversity (epsilon 0.1) over Attention baselineOurs (Tdiv, 0.1) · MS COCO testSPICE: 19.9 -> 21.0Improved
CIDEr optimization with reinforcement learningCIDEr-RL · MS COCO testDist-3: 4972 -> 3694Hurt

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