AI Paper Atlas

Research map / Vision-language and multimodal / CLIP and zero-shot

ArtVLM: Attribute Recognition Through Vision-Based Prefix Language Modeling

William Yicheng Zhu, Keren Ye, Junjie Ke, Jiahui Yu, Leonidas Guibas, Peyman Milanfar, Feng Yang

ECCV 2024 · Vision-language and multimodal › CLIP and zero-shot · approach: Open-vocabulary alignment from VLMs · 2 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 generative retrieval instead of contrastive retrievalCoCa Base · VAW zero-shotaverage rank: 95.1 -> 56.0Improved
use generative retrieval instead of contrastive retrieval after finetuningCoCa Base · VAW finetunedaverage rank: 12.2 -> 10.6Improved
add object hint to contrastive retrieval sentence (ablation)Contrastive retrieval · VAW zero-shotaverage rank: 95.1 -> 149.8Hurt
use object-conditioned sentence {A}{O} instead of attribute-only {A} in generative retrieval (ablation)Generative retrieval · VAW zero-shotaverage rank: 82.1 -> 63.9Improved
use hybrid template {A}{O} is {A} instead of {O} is {A} (ablation)Generative retrieval · VAW zero-shotaverage rank: 61.9 -> 56.0Improved
use hybrid template {A}{O} is {A} on object recognition instead of {A}{O} (ablation)Generative retrieval · VGARank-Object zero-shotrank: 5.8 -> 6.4Hurt
use hybrid template on attribute recognition (ablation)Generative retrieval · VGARank-Attribute zero-shotrank: 13.0 -> 12.0Improved
use generative retrieval instead of contrastive on VGARank-ObjectCoCa Base · VGARank-Object zero-shotrank: 5.9 -> 5.8No clear effect
use generative retrieval with CoCa instead of TAP with in-domain pretrainingVAW · VAW SOTA comparisonoverall mAP: 73.4 -> 72.0Hurt
use generative retrieval instead of SCoNE on rare attributesVAW · VAW Tail mAPmAP (%): 48.0 -> 60.6Improved

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