Related papers: Deep Learning for Line Intensity Mapping Observati…
To implement autonomous driving, one essential step is to model the vehicle environment based on the sensor inputs. Radars, with their well-known advantages, became a popular option to infer the occupancy state of grid cells surrounding the…
Deep neural networks have proven to be highly effective when large amounts of data with clean labels are available. However, their performance degrades when training data contains noisy labels, leading to poor generalization on the test…
We introduce a novel method for training machine learning models in the presence of noisy labels, which are prevalent in domains such as medical diagnosis and autonomous driving and have the potential to degrade a model's generalization…
The problem of learning structural equation models (SEMs) from data is a fundamental problem in causal inference. We develop a new algorithm --- which is computationally and statistically efficient and works in the high-dimensional regime…
We use Generative Adversarial Networks (GANs) to design a class conditional label noise (CCN) robust scheme for binary classification. It first generates a set of correctly labelled data points from noisy labelled data and 0.1% or 1% clean…
The matter density is an important knowledge for today cosmology as many phenomena are linked to matter fluctuations. However, this density is not directly available, but estimated through lensing maps or galaxy surveys. In this article, we…
Real-world image noise removal is a long-standing yet very challenging task in computer vision. The success of deep neural network in denoising stimulates the research of noise generation, aiming at synthesizing more clean-noisy image pairs…
Deep neural networks (DNNs) have achieved great success in a wide variety of medical image analysis tasks. However, these achievements indispensably rely on the accurately-annotated datasets. If with the noisy-labeled images, the training…
Noisy labels are commonly found in real-world data, which cause performance degradation of deep neural networks. Cleaning data manually is labour-intensive and time-consuming. Previous research mostly focuses on enhancing classification…
We present a model for the power spectrum of the density field of galaxies weighted by their star formation rate. This weighting is relevant in line-intensity mapping (LIM) when the observed line luminosity is strongly correlated with star…
Label noise may affect the generalization of classifiers, and the effective learning of main patterns from samples with noisy labels is an important challenge. Recent studies have shown that deep neural networks tend to prioritize the…
We discuss methods for performing weak lensing using radio observations to recover information about the intrinsic structural properties of the source galaxies. Radio surveys provide unique information that can benefit weak lensing studies,…
Recent studies indicate that deep neural networks degrade in generalization performance under noisy supervision. Existing methods focus on isolating clean subsets or correcting noisy labels, facing limitations such as high computational…
Explainability algorithms such as LIME have enabled machine learning systems to adopt transparency and fairness, which are important qualities in commercial use cases. However, recent work has shown that LIME's naive sampling strategy can…
Weak lensing surveys have become a powerful tool for mapping mass distributions and constraining the expansion history of our Universe, but continuum surveys must average over a large number of galaxies to average down the ellipticity noise…
This work proposes a Stochastic Variational Deep Kernel Learning method for the data-driven discovery of low-dimensional dynamical models from high-dimensional noisy data. The framework is composed of an encoder that compresses…
Understanding the formation and evolution of the Universe is crucial for cosmological studies, and the line intensity mapping provides a powerful tool for this kind of study. We propose to make use of multipole moments of redshift-space…
Intensity mapping has emerged as a promising tool to probe the three-dimensional structure of the universe. The traditional approach of galaxy redshift surveys is based on individual galaxy detection, typically performed by thresholding and…
Deep Learning (DL) and Deep Neural Networks (DNNs) are widely used in various domains. However, adversarial attacks can easily mislead a neural network and lead to wrong decisions. Defense mechanisms are highly preferred in safety-critical…
We explore the possible application of linear covariance-based (LCB) filtering to line-intensity mapping (LIM) signal reconstructions. Originally introduced for reconstruction of the integrated Sachs-Wolfe effect in the cosmic microwave…