相关论文: Rethinking InfoNCE: How Many Negative Samples Do Y…

Through minimization of an appropriate loss function such as the InfoNCE loss, contrastive learning (CL) learns a useful representation function by pulling positive samples close to each other while pushing negative samples far apart in the…

Measurement samples are often taken in various monitoring applications. To reduce the sensing cost, it is desirable to achieve better sensing quality while using fewer samples. Compressive Sensing (CS) technique finds its role when the…

Learning with noisy labels has gained increasing attention because the inevitable imperfect labels in real-world scenarios can substantially hurt the deep model performance. Recent studies tend to regard low-loss samples as clean ones and…

This work proposes a new loss function targeting classification problems, utilizing a source of information overlooked by cross entropy loss. First, we derive a series of the tightest upper and lower bounds for the probability of a random…

Model calibration aims to align confidence with prediction correctness. The Cross-Entropy (CE) loss is widely used for calibrator training, which enforces the model to increase confidence on the ground truth class. However, we find the CE…

There are many models, often called unnormalized models, whose normalizing constants are not calculated in closed form. Maximum likelihood estimation is not directly applicable to unnormalized models. Score matching, contrastive divergence…

Learning to rank has recently emerged as an attractive technique to train deep convolutional neural networks for various computer vision tasks. Pairwise ranking, in particular, has been successful in multi-label image classification,…

Hard negatives play a critical role in training and fine-tuning dense retrieval models, as they are semantically similar to positive documents yet non-relevant, and correctly distinguishing them is essential for improving retrieval…

Prior theory work on Contrastive Learning via the InfoNCE loss showed that, under certain assumptions, the learned representations recover the ground-truth latent factors. We argue that these theories overlook crucial aspects of how CL is…

We review the current schemes of text-image matching models and propose improvements for both training and inference. First, we empirically show limitations of two popular loss (sum and max-margin loss) widely used in training text-image…

A common problem in contextual preference ranking is that a single preferred action is compared against several choices, thereby blowing up the complexity and skewing the preference distribution. In this work, we show how one can solve this…

Learning a parametric model of a data distribution is a well-known statistical problem that has seen renewed interest as it is brought to scale in deep learning. Framing the problem as a self-supervised task, where data samples are…

We consider training decision trees using noisily labeled data, focusing on loss functions that can lead to robust learning algorithms. Our contributions are threefold. First, we offer novel theoretical insights on the robustness of many…

Knowledge graph embedding (KGE) models encode the structural information of knowledge graphs to predicting new links. Effective training of these models requires distinguishing between positive and negative samples with high precision.…

A common approach in positive-unlabeled learning is to train a classification model between labeled and unlabeled data. This strategy is in fact known to give an optimal classifier under mild conditions; however, it results in biased…

Current debiasing approaches often result a degradation in model capabilities such as factual accuracy and knowledge retention. Through systematic evaluation across multiple benchmarks, we demonstrate that existing debiasing methods face…

Conventional semi-supervised contrastive learning methods assign pseudo-labels only to samples whose highest predicted class probability exceeds a predefined threshold, and then perform supervised contrastive learning using those selected…

We consider the problem of retrieving the most relevant labels for a given input when the size of the output space is very large. Retrieval methods are modeled as set-valued classifiers which output a small set of classes for each input,…

Semi-supervised learning (SSL) constructs classifiers using both labelled and unlabelled data. It leverages information from labelled samples, whose acquisition is often costly or labour-intensive, together with unlabelled data to enhance…

In Multi-Label Learning (MLL), it is extremely challenging to accurately annotate every appearing object due to expensive costs and limited knowledge. When facing such a challenge, a more practical and cheaper alternative should be Single…