RT Conference Proceedings
T1 Moving object detection in noisy video sequences using deep convolutional disentangled representations.
A1 García-González, Jorge
A1 Luque-Baena, Rafael Marcos
A1 Ortiz-de-Lazcano-Lobato, Juan Miguel
A1 López-Rubio, Ezequiel
K1 Procesado de imágenes
AB Noise robustness is crucial when approaching a moving de-tection problem since image noise is easily mistaken formovement. In order to deal with the noise, deep denoisingautoencoders are commonly proposed to be applied on imagepatches with an inherent disadvantage with respect to thesegmentation resolution. In this work, a fully convolutionalautoencoder-based moving detection model is proposed inorder to deal with noise with no patch extraction required.Different autoencoder structures and training strategies arealso tested to get insights into the best network design ap-proach.
YR 2022
FD 2022
LK https://hdl.handle.net/10630/25286
UL https://hdl.handle.net/10630/25286
LA eng
NO Universidad de Málaga. Campus de Excelencia Internacional Andalucía Tech.
DS RIUMA. Repositorio Institucional de la Universidad de Málaga
RD 20 ene 2026