Search Results for author:
Found 5 papers, 0 papers with code
Learning by Hallucinating: Vision-Language Pre-training with Weak Supervision
Moreover, in other V-L downstream tasks considered, our WFH models are on par with models trained with paired V-L data, revealing the utility of unpaired data.
Learning to hash with semantic similarity metrics and empirical KL divergence
We address (ii) via a differentiable estimate of the KL divergence between network outputs and a binary target distribution, resulting in minimal information loss when the features are rounded to binary.
SHREWD: Semantic Hierarchy-based Relational Embeddings for Weakly-supervised Deep Hashing
Using class labels to represent class similarity is a typical approach to training deep hashing systems for retrieval; samples from the same or different classes take binary 1 or 0 similarity values.
SHREWD: Semantic Hierarchy Based Relational Embeddings For Weakly-Supervised Deep Hashing
Using class labels to represent class similarity is a typical approach to training deep hashing systems for retrieval; samples from the same or different classes take binary 1 or 0 similarity values.
On the exact relationship between the denoising function and the data distribution
We prove an exact relationship between the optimal denoising function and the data distribution in the case of additive Gaussian noise, showing that denoising implicitly models the structure of data allowing it to be exploited in the unsupervised learning of representations.
