Related papers: A Sub-Millisecond Fourier and Wavelet Based Model …

The Near-Earth Object Wide-field Infrared Explorer (NEOWISE) mission provides a decade of all-sky time-series data at 3.4 and 4.6um and an unprecedented opportunity for the discovery and characterization of variable objects. This paper…

Building on prior FAST targeted and blind SETI campaigns toward 33 exoplanet systems, we introduce a wavelet-integrated search pipeline for narrowband technosignature candidates in radio dynamic spectra. At its core, the pipeline uses a…

The WISE and NEOWISE missions have provided the only mid-infrared all-sky time-domain data, opening a unique observational window for variability studies. Yet, a comprehensive and systematic catalog of mid-infrared variable sources has…

[abridged] In large-scale time-domain surveys, the processing of data, from procurement up to the detection of sources, is generally automated. One of the main challenges is contamination by artifacts, especially in regions of strong…

The heavy-tailed nature of precipitation intensity impedes precise precipitation nowcasting. Standard models that optimize pixel-wise losses are prone to regression-to-the-mean bias, which blurs extreme values. Existing Fourier-based…

Transformer-based image denoising methods have achieved encouraging results in the past year. However, it must uses linear operations to model long-range dependencies, which greatly increases model inference time and consumes GPU storage…

Deep image registration has demonstrated exceptional accuracy and fast inference. Recent advances have adopted either multiple cascades or pyramid architectures to estimate dense deformation fields in a coarse-to-fine manner. However, due…

High-frequency features are critical in multiscale phenomena such as turbulent flows and phase transitions, since they encode essential physical information. The recently proposed Wavelet Neural Operator (WNO) utilizes wavelets'…

Video Variational Autoencoder (VAE) encodes videos into a low-dimensional latent space, becoming a key component of most Latent Video Diffusion Models (LVDMs) to reduce model training costs. However, as the resolution and duration of…

In this paper we propose a new model-based unsupervised learning method, called VarNet, for the solution of partial differential equations (PDEs) using deep neural networks (NNs). Particularly, we propose a novel loss function that relies…

Express Wavenet is an improved optical diffractive neural network. At each layer, it uses wavelet-like pattern to modulate the phase of optical waves. For input image with n2 pixels, express wavenet reduce parameter number from O(n2) to…

Time series are ubiquitous in many applications that involve forecasting, classification and causal inference tasks, such as healthcare, finance, audio signal processing and climate sciences. Still, large, high-quality time series datasets…

We present FourierNet, a single shot, anchor-free, fully convolutional instance segmentation method that predicts a shape vector. Consequently, this shape vector is converted into the masks' contour points using a fast numerical transform.…

The Wide-field Infrared Survey Explorer Preliminary Data Release Source Catalog contains over 257 million objects. We describe the method used to flag variable source candidates in the Catalog. Using a method based on the chi- square of…

Objective. Limited access to breast cancer diagnosis globally leads to delayed treatment. Ultrasound, an effective yet underutilized method, requires specialized training for sonographers, which hinders its widespread use. Approach. Volume…

The wavelet transform has been used for numerous studies in astrophysics, including signal--noise periodicity and decomposition as well as the signature of differential rotation in stellar light curves. In the present work, we apply the…

We describe a methodology to classify periodic variable stars identified using photometric time-series measurements constructed from the Wide-field Infrared Survey Explorer (WISE) full-mission single-exposure Source Databases. This will…

While diffusion-based generative models have made significant strides in visual content creation, conventional approaches face computational challenges, especially for high-resolution images, as they denoise the entire image from noisy…

Visual Auto-Regressive (VAR) models significantly reduce inference steps through the "next-scale" prediction paradigm. However, progressive multi-scale generation incurs substantial memory overhead due to cumulative KV caching, limiting…

Stellar photometric variability and instrumental effects, like cosmic ray hits, data discontinuities, data leaks, instrument aging etc. cause difficulties in the characterization of exoplanets and have an impact on the accuracy and…