AI Paper Atlas

Research map / Vision-language and multimodal / Cross-modal learning

Unpaired Image-to-Speech Synthesis With Multimodal Information Bottleneck

Shuang Ma, Daniel McDuff, Yale Song

ICCV 2019 · Vision-language and multimodal › Cross-modal learning · approach: Contrastive vision-language alignment · 28 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 memory fusion module M (ablation)Ours · image-to-text (COCO)BLEU-1: 65.2 -> 74.1Improved
add the memory fusion module M (ablation)Ours · speech-to-textWER: 6.99 -> 3.88Improved
use memory module M instead of two FC layers of similar size (ablation)Ours · image-to-textBLEU-1: 65.9 -> 74.1Improved
add the modality transformer T (ablation)Ours · image-to-textBLEU-1: 68.6 -> 74.1Improved
add the modality transformer T (ablation)Ours · speech-to-textWER: 6.01 -> 3.88Improved
add the adversarial loss L_adv to train T (ablation)Ours · image-to-textBLEU-1: 69.8 -> 74.1Improved
add the adversarial loss L_adv to train T (ablation)Ours · speech-to-textWER: 5.87 -> 3.88Improved
use end-to-end skip-modal model instead of piecewise image-to-text then text-to-speechimage-to-speech · COCO + EMT-4WER: 4.1 -> 3.9Improved
use skip-modal model instead of piecewise baselineimage-to-speech · COCO + EMT-4BLEU-4: 30.1 -> 29.9No clear effect
use CycleGAN cross-domain translation for image-to-speechimage-to-speech · COCO + EMT-4BLEU-1: 69.2 -> 26.2Hurt
swap the 3-layer CNN encoder for UD encoderOurs · COCO image captioningBLEU-4: 30.6 -> 37.1Improved
mix both datasets in each batch instead of alternating (ablation)Ours · speech-to-textWER: 3.88 -> 3.76Improved

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