Convolutional LSTM

Convolutional LSTMs (ConvLSTMs) combine the spatial feature extraction capabilities of convolutional neural networks (CNNs) with the temporal processing power of long short-term memory (LSTM) networks to analyze spatiotemporal data. Current research focuses on applying ConvLSTMs to diverse prediction tasks, including weather forecasting (using radar and satellite imagery), groundwater level prediction, and human activity recognition from video, often within autoencoder frameworks. These models demonstrate significant potential for improving accuracy in various fields by leveraging both spatial and temporal dependencies within data, leading to more robust and informative predictions. However, comparisons with alternative architectures, such as transformers and feedforward CNNs, highlight ongoing efforts to optimize model performance and efficiency for specific applications.

Papers

November 14, 2024

Self-Supervised Radio Pre-training: Toward Foundational Models for Spectrogram Learning
Ahmed Aboulfotouh, Ashkan Eshaghbeigi, Dimitrios Karslidis, Hatem Abou-Zeid
Deep Learning Self Supervised Learning Foundational Model Convolutional LSTM Radio Data

September 20, 2024

Time Distributed Deep Learning models for Purely Exogenous Forecasting. Application to Water Table Depth Prediction using Weather Image Time Series
Matteo Salis, Abdourrahmane M. Atto, Stefano Ferraris, Rosa Meo
Deep Learning Model Time Series Time Matter LSTM Model Groundwater Flow Convolutional LSTM

December 2, 2023

Short-term Precipitation Forecasting in The Netherlands: An Application of Convolutional LSTM neural networks to weather radar data
Petros Demetrakopoulos
Neural Network Long Short Term Memory Radar Data ConvLSTM Network Convolutional LSTM

October 3, 2023

Improving style transfer in dynamic contrast enhanced MRI using a spatio-temporal approach
Adam G. Tattersall, Keith A. Goatman, Lucy E. Kershaw, Scott I. K. Semple, Sonia Dahdouh
Supervised Autoencoder Style Transfer Spatio Temporal Magnetic Resonance Imaging Contrast Enhancement Color Contrast Dce Mri Convolutional LSTM

April 18, 2023

Human activity recognition using deep learning approaches and single frame cnn and convolutional lstm
Sheryl Mathew, Annapoorani Subramanian, Pooja, Balamurugan MS, Manoj Kumar Rajagopal
Deep Learning Long Short Term Memory Activity Recognition Action Recognition Benchmark Single Convolutional Convolutional LSTM

October 22, 2022

Baby Physical Safety Monitoring in Smart Home Using Action Recognition System
Victor Adewopo, Nelly Elsayed, Kelly Anderson
Action Recognition Convolutional LSTM Action Recognition System

June 26, 2022

Video Anomaly Detection via Prediction Network with Enhanced Spatio-Temporal Memory Exchange
Guodong Shen, Yuqi Ouyang, Victor Sanchez
Long Short Term Memory Video Anomaly Detection Temporal Memory Convolutional LSTM Prediction Network Temporal Regularity

March 8, 2022

Visual anomaly detection in video by variational autoencoder
Faraz Waseem, Rafael Perez Martinez, Chris Wu
Supervised Autoencoder Variational Autoencoder Source Video Video Anomaly Detection Visual Anomaly Detection Convolutional LSTM LSTM Autoencoder

February 23, 2022

A Bayesian Deep Learning Approach to Near-Term Climate Prediction
Xihaier Luo, Balasubramanya T. Nadiga, Yihui Ren, Ji Hwan Park, Wei Xu, Shinjae Yoo
Bayesian Deep Learning Ensemble Weather Climate Prediction Convolutional LSTM Decadal Temperature Prediction

December 7, 2021

Emulating Spatio-Temporal Realizations of Three-Dimensional Isotropic Turbulence via Deep Sequence Learning Models
Mohammadreza Momenifar, Enmao Diao, Vahid Tarokh, Andrew D. Bragg
Deep Learning Framework Turbulent Flow Deep Sequence Convolutional LSTM Gradient Tensor

November 18, 2021

MS-nowcasting: Operational Precipitation Nowcasting with Convolutional LSTMs at Microsoft Weather
Sylwester Klocek, Haiyu Dong, Matthew Dixon, Panashe Kanengoni, Najeeb Kazmi, Pete Luferenko, Zhongjian Lv, Shikhar Sharma, Jonathan Weyn, Siqi Xiang
Deep Learning Model Long Short Term Memory Data Driven Nowcasting Precipitation Nowcasting Convolutional LSTM Radar Reflectivity

Convolutional LSTM

Papers

Self-Supervised Radio Pre-training: Toward Foundational Models for Spectrogram Learning

Time Distributed Deep Learning models for Purely Exogenous Forecasting. Application to Water Table Depth Prediction using Weather Image Time Series

Short-term Precipitation Forecasting in The Netherlands: An Application of Convolutional LSTM neural networks to weather radar data

Improving style transfer in dynamic contrast enhanced MRI using a spatio-temporal approach

Human activity recognition using deep learning approaches and single frame cnn and convolutional lstm

Baby Physical Safety Monitoring in Smart Home Using Action Recognition System

Video Anomaly Detection via Prediction Network with Enhanced Spatio-Temporal Memory Exchange

Visual anomaly detection in video by variational autoencoder

A Bayesian Deep Learning Approach to Near-Term Climate Prediction

Emulating Spatio-Temporal Realizations of Three-Dimensional Isotropic Turbulence via Deep Sequence Learning Models

MS-nowcasting: Operational Precipitation Nowcasting with Convolutional LSTMs at Microsoft Weather