Video Prediction on V-MNIST
25.1MSEPredRNN + MotionRNN
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| PredRNN + MotionRNNbackbone=PredRNN, framework=MotionRNN2021.03 | 25.1 | 92 | 24 | 27.7 | |
| MIM + MotionRNNbackbone=MIM, framework=MotionRNN2021.03 | 28.9 | 90.6 | 23.1 | 30.9 | |
| MIM2021.03 | 34.6 | 88.8 | 22.3 | 34.6 | |
| PredRNN2021.03 | 35.6 | 89.1 | 22.1 | 34.7 | |
| ConvLSTM + MotionRNNbackbone=ConvLSTM, framework=MotionRNN2021.03 | 44.4 | 86.1 | 20.9 | 40.3 | |
| ConvLSTM2021.03 | 47 | 84.5 | 20.6 | 41.8 | |
| E3D-LSTM + MotionRNNbackbone=E3D-LSTM, framework=MotionRNN2021.03 | 52.8 | 86.7 | 20.3 | 42.4 | |
| E3D-LSTM2021.03 | 57.6 | 85.2 | 19.7 | 44.6 | |
| Conv-TT-LSTM2021.03 | 71.1 | 74.4 | 18.4 | 53.6 | |
| TrajGRU2021.03 | 109 | 51.5 | 15.9 | 69.3 |