16 datasets found

The sub-sentence encoder is a contrastive learning framework for learning contextual embeddings for semantic units on the sub-sentence level.

Contrastive learning is one of the most successful paradigms for self-supervised learning (SSL). In a principled way, it considers two augmented views of the same image as...

Contrastive Visual-Linguistic Pretraining (CVLP) is a novel approach to visual-linguistic pretraining that solves the domain bias and noisy label problems encountered with...

The dataset used in the paper is a collection of pairs of observations (xi, ˜xi) from two modalities, where xi ∈ Rd1 and ˜xi ∈ Rd2. The dataset is used to evaluate the...

Self-supervised contrastive learning models, such as CLIP, have set new benchmarks for vision-language models in many downstream tasks. However, their dependency on rigid...

Adaptive Multi-head Contrastive Learning (AMCL) framework, which tackles intra- and inter-sample similarity and adaptive temperature mechanism re-weighting each similarity pair.

Self-supervised learning framework that maintains the relational consistency between instances under different augmentations.

The existing contrastive learning methods widely adopt one-hot instance discrimination as pretext task for self-supervised learning, which inevitably neglects rich...

Contrastive learning for compact single image dehazing

Contrastive multiview coding.

A simple framework for contrastive learning of visual representations.

Contrastive self-supervised learning (CSL) with a prototypical regularization has been introduced in learning meaningful representations for downstream tasks that require strong...

A simple method that minimizes computational resource footprint for CLIP (Contrastive Language Image Pretraining).

The data encountered by Session-based Recommendation System(SBRS) is typically highly sparse, which also serves as one of the bottlenecks limiting the accuracy of recommendations.

The CLIP dataset is used to train a contrastive learning model.

We propose DisCo-CLIP, a distributed memory-efficient CLIP training approach, to reduce the memory consump- tion of contrastive loss when training contrastive learning models.