AI Paper Atlas

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

I2DFormer: Learning Image to Document Attention for Zero-Shot Image Classification

Muhammad Ferjad Naeem, Yongqin Xian, Luc Van Gool, Federico Tombari

NeurIPS 2022 · Vision-language and multimodal › CLIP and zero-shot · approach: Attribute-based zero-shot learning · 21 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 I2D Attention to I2DGlobal (ablation)I2DFormer · AWA2, CUB, and FLOZSL T1: 69.4 -> 76.4; 37.2 -> 45.4; 37.2 -> 40.0Improved
replace I2D Attention with FILIP (ablation)I2DGlobal · AWA2, CUB, and FLOZSL T1: 69.4 -> 67.3; 37.2 -> 35.7; 37.2 -> 38.3Hurt
replace I2D Attention with ViLBERT cross-modal and self-attention (ablation)I2DGlobal · AWA2, CUB, and FLOZSL T1: 69.4 -> 75.0; 37.2 -> 29.9; 37.2 -> 21.3Mixed
use LongFormer instead of GloVe input embeddings (ablation)I2DFormer · AWA2, CUB, and FLOZSL T1: 76.4 -> 51.5; 45.4 -> 34.3; 40.0 -> 21.8Hurt
use MPNet instead of GloVe input embeddings (ablation)I2DFormer · AWA2, CUB, and FLOZSL T1: 76.4 -> 65.5; 45.4 -> 36.1; 40.0 -> 28.3Hurt
use Word2Vec instead of GloVe input embeddings (ablation)I2DFormer · AWA2, CUB, and FLOZSL T1: 76.4 -> 74.0; 45.4 -> 43.8; 40.0 -> 37.8Hurt
learn image-document semantic embeddings from documents rather than class-name GloVe embeddingsI2DFormer global module · AWA2, CUB, and FLOZSL T1: 52.1 -> 61.6; 20.4 -> 29.0; 21.6 -> 25.8Improved
use I2DEmb instead of GloVe document embeddingssemantic embedding methods · AWA2, CUB, and FLOZSL T1: 61.6 -> 76.4; 29.0 -> 45.4; 25.8 -> 40.0Improved
use I2DFormer rather than the strongest unsupervised embedding baselineI2DFormer · AWA2, CUB, and FLOGZSL H: 66.0 -> 71.5; 37.3 -> 43.8; 44.3 -> 51.5Improved

Later papers that compared with this method

1 later accepted paper reports I2DFormer+ as a baseline in their results tables; 0 of them report a better number on most of the results they share with it.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

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