Devdiscourse

Fast and Accurate Water Leak Detection in Distribution Networks Using Deep Anomaly Models

Researchers from Politecnico di Milano propose a data-driven water leak detection method that treats leaks as anomalies in ...
GitHub

CONVOLUTIONAL-AUTOENCODER-FOR-IMAGE-DENOISING

This code implements a Denoising Autoencoder using PyTorch to clean noisy images from the MNIST dataset. It uses a convolutional neural network architecture, where the encoder compresses the input ...
note

機械学習モデルの実装：AutoEncoder（Pytorch）

本記事ではAutoEncoderをPytorchで実装します。ライブラリ紹介ではなく実装がメインのため学習シリーズに近い内容となります。 AutoEncoderの特徴は下記の通りです。 情報を圧縮することで軽い動作で使用できる。 ー＞本記事では784次元の画像を200次元（潜在 ...
IEEE

A Lightweight Convolutional Autoencoder for Moiré Pattern Removal with Enhanced Text ...

Abstract: Moiré patterns, caused by aliasing in screen-captured images, degrade visual quality and challenge text readability in applications like document digitization. This paper proposes a ...
IEEE

Enhanced demodulator for 5G NTN using spatio-temporal attention convolutional autoencoder ...

Abstract: This paper presents a novel approach to overcoming challenges in 5G and 6G mobile satellite systems (MSS) in Low Earth Orbit (LEO), focusing on Non-Line-of-Sight (NLOS) issues in 5G ...