List your company

Precipitation nowcasting on Radar Echo dataset (test)

43.1CSI@30MIM + MotionRNN
24.3829.2434.138.96Mar 3, 2021

Evaluation results

MethodLinks
MIM + MotionRNNbackbone=MIM, MotionRNN=enabled2021.0343.1——————40.334.3
MIMbackbone=MIM2021.0342.9——————39.931.7
PredRNN + MotionRNNbackbone=PredRNN, MotionRNN=enabled2021.0342.4——————39.134.5
ConvLSTM + MotionRNNbackbone=ConvLSTM, MotionRNN=enabled2021.0341.1——————38.735
PredRNNbackbone=PredRNN2021.0340.1——————37.830.6
ConvLSTMbackbone=ConvLSTM2021.0338.5——————3631.5
E3D-LSTM + MotionRNNbackbone=E3D-LSTM, MotionRNN=enabled2021.0337.9——————36.633.3
E3D-LSTMbackbone=E3D-LSTM2021.0335——————30.925.4
TrajGRUbackbone=TrajGRU2021.0325.1——————20.115
Conv-LSTMInput window=5 images, Prediction window=15 images, Image resolution=100x1002018.06—1.4257.719.5660.908———
Conv-LSTM-BMSInput window=5 images, Prediction window=15 images, Image resolution=100x100, Number of samples (T)=12018.06—1.1636716.373.40.918132.52——
Conv-LSTM-CVAEInput window=5 images, Prediction window=15 images, Image resolution=100x100, Number of samples (T)=12018.06—1.25965.115.570.10.91132.78——
ConvLSTMKernel Size=3x3, Architecture=3x3-64-3x3-64, Prediction Steps=152015.06—1.4257.719.5660.908———
FC-LSTMArchitecture=2000-2000, Prediction Steps=152015.06—1.86528.633.535.10.774———
Rover1Prediction Steps=152015.06—1.71251.630.863.60.843———
Rover2Prediction Steps=152015.06—1.68452.230.164.20.85———
Rover3Prediction Steps=152015.06—1.68552.230.164.20.849———