Embedding Geometries of Contrastive Language-Image Pre-Training

Jason Chuan-Chih Chou, Nahid Alam
9/19/2024
cs.LGcs.CLcs.CV

No confident code implementation found for this paper. 2 possible matches can be shown below.

Abstract

Since the publication of CLIP, the approach of using InfoNCE loss for contrastive pre-training has become widely popular for bridging two or more modalities. Despite its wide adoption, CLIP's original design choices of L2 normalization and cosine similarity logit have rarely been revisited. We have systematically experimented with alternative geometries and softmax logits for language-image pre-training and identified that variants with intuitive Euclidean geometry, Euclidean CLIP (EuCLIP), match or exceed the performance of CLIP and support hierarchical relationships at least as well as more complicated hyperbolic alternative.

View on arXivView PDF

Code Implementations

No confident code match yet

We couldn't find an author-owned or strongly-evidenced community implementation for this paper. 2 weaker matches are hidden by default — verify before relying on them.

No code implementations found yet.

Know of an implementation? Let us know in the comments below!

Similar Papers

Data-Efficient Contrastive Language-Image Pretraining: Prioritizing Data Quality over Quantity

84% match1 repo102024

Is CLIP ideal? No. Can we fix it? Yes!

84% match1 repo562025

CLIP-Map: Structured Matrix Mapping for Parameter-Efficient CLIP Compression

83% match2026

Contrastive Localized Language-Image Pre-Training

83% match1 repo1852024

Improving CLIP Training with Language Rewrites

82% match1 repo2912023

SPACE-CLIP: Spatial Perception via Adaptive CLIP Embeddings for Monocular Depth Estimation

82% match1 repo22026

Cite this paper

@article{chou2024embedding,
  title  = {Embedding Geometries of Contrastive Language-Image Pre-Training},
  author = {Jason Chuan-Chih Chou and Nahid Alam},
  year   = {2024},
  eprint = {2409.13079},
  archivePrefix = {arXiv},
  url    = {http://arxiv.org/abs/2409.13079v1}
}

Add a badge to your README

Shows whether official code is linked for this paper.

Code of Paper badge
[![Code of Paper](https://codeofpaper.com/badge/2409.13079.svg)](https://codeofpaper.com/paper/2409.13079)

Discussion