Study on the Influence of Attention Mechanism in Large-Scale Sea Surface Temperature Prediction Based on Temporal Convolutional Network

Feng, Yuan; Sun, Tianying; Li, Chen

doi:10.1007/978-3-030-89814-4_53

Cited by 1 publication

(2 citation statements)

References 9 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Recently, several spatiotemporal attention mechanisms have since been developed to classify or predict sea surface data such as SSH, SST. Feng et al applied time attention mechanism and temporal convolutional network to construct fullfeature and partial-feature prediction models for large-scale SST data, achieving similar accuracy with less data [60]. Based on spatial attention mechanism, Ren et al proposed a dualattention U-Net for pixel-level segmentation of sea ice and open water, with better classification results than the original U-Net [30].…”

Section: B Prediction Network With Attention Mechanismsmentioning

confidence: 99%

See 1 more Smart Citation

Spatio–Temporal Attention-Based Deep Learning Framework for Mesoscale Eddy Trajectory Prediction

Wang

Liu

Wang

et al. 2022

IEEE J. Sel. Top. Appl. Earth Observations Remote Sensing

View full text Add to dashboard Cite

Accurate prediction of mesoscale eddy trajectories requires efficient models with large-size available data instances to capture the main eddy characteristics. However, there is a lack of salient attention mechanisms that can recognize the demand of extracting the aggregated features over multi-dimensional eddy data sources. Additionally, deep learning techniques are very important for eddy trajectory prediction to dynamically capture properties of mesoscale eddies in the South China Sea (SCS). In this paper, we propose a spatio-temporal attention-based deep learning framework that can orchestrate heterogeneous data integration and propagation trajectory forecast together. It consists of a novel autoencoder equipped with channel and spatial attention mechanisms (CSA-Encoder), and a gated recurrent unit (GRU) network with temporal attention layer (TA-GRU). CSA-Encoder compresses stereoscopic eddy data with convolutional layers and generates the small-scale and high-quality dataset as the input of TA-GRU. The finer-grained TA-GRU method is extended to accurately predict eddy trajectories with more valuable imagery information so that the temporal attention mechanism can automatically select relevant regions within the next 14 days. Our cross-validation results demonstrate that our framework averagely achieves a lower distance error (9 km) and 54% performance improvement over the baseline GRU technique in the next one day, and outperforms two state-of-theart techniques of long short-term memory (LSTM) and recurrent neural network (RNN) by 54.9% and 65.6%, respectively.

show abstract

Section: B Prediction Network With Attention Mechanismsmentioning

confidence: 99%

“…Furthermore, to probe the effect of compressed data layer number on CSA-Encoder's MSE loss, the results with different compressed layer numbers (Compressed Data Layers: 1,5,10,20,30,40,50,60)…”

Section: B Comparison Of Mse Loss In Csa-encodermentioning

confidence: 99%