AI Paper Atlas

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

Learnable PINs: Cross-Modal Embeddings for Person Identity

Arsha Nagrani, Samuel Albanie, Andrew Zisserman

ECCV 2018 · Vision-language and multimodal › Cross-modal learning · approach: Probabilistic embeddings for cross-modal retrieval · 157 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
initialise subnetworks with pretrained single-modality identity weightsJoint embedding · seen-heard verificationAUC (%): 73.8 -> 87.0Improved
initialise subnetworks with pretrained single-modality identity weightsJoint embedding · unseen-unheard verificationAUC (%): 63.5 -> 78.5Improved
initialise subnetworks with pretrained single-modality identity weightsJoint embedding · unseen-unheard verificationEER (%): 39.2 -> 29.6Improved
train from scratch with no identity labels versus random weightsJoint embedding · unseen-unheard verificationAUC (%): 50.1 -> 63.5Improved
choose negative pairs with same genderJoint embedding (pretrained) · unseen-unheard verificationAUC (%): 78.5 -> 61.1Hurt
choose negative pairs with same nationalityJoint embedding (pretrained) · unseen-unheard verificationAUC (%): 78.5 -> 77.2Hurt
choose negative pairs with same ageJoint embedding (pretrained) · unseen-unheard verificationAUC (%): 78.5 -> 74.9Hurt
sample positive pairs from the same video instead of across videosJoint embedding (scratch) · seen-heard verificationAUC (%): 73.8 -> 74.5Improved
sample positive pairs from the same face-trackJoint embedding (scratch) · seen-heard verificationAUC (%): 74.5 -> 74.6No clear effect
add voice embeddings via max fusion to face-only retrievalFace-only retrieval · Sherlock one-shot retrieval (Sherlock)AUC (%): 35.0 -> 37.5Improved
add voice embeddings via max fusion to face-only retrievalFace-only retrieval · Sherlock one-shot retrieval (John)AUC (%): 44.6 -> 45.4Improved
freeze pretrained face subnetwork as teacher and train voice from scratch (ablation)Joint embedding · verificationaccuracyHurt

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