REVIEW 1 cited by
Embedding Geometries of Contrastive Language-Image Pre-Training
Not yet reviewed by Pith; the record is open.
This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.
SPECIMEN: schema-true, not a live event
T0 review · schema-true
One-sentence machine reading of the paper's core claim.
pith:XXXXXXXX · record.json · timestamp
Embedding Geometries of Contrastive Language-Image Pre-Training
read the original 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.
Forward citations
Cited by 1 Pith paper
-
Language-based Color ISP Tuning
Language-described color styles can be applied to photos by optimizing a small camera color matrix with gradient descent against a vision-language model's similarity score.
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.