Forecast-MAE: Self-supervised Pre-training for Motion Forecasting with Masked Autoencoders.
Jie ChengXiaodong MeiMing LiuPublished in: ICCV (2023)
Keyphrases
- short term
- forecasting model
- image sequences
- motion estimation
- forecasting accuracy
- early warning
- short term load forecasting
- failure prediction
- optical flow
- long term
- grey theory
- feedforward neural networks
- training process
- motion analysis
- motion model
- space time
- arima model
- exponential smoothing
- camera motion
- financial forecasting
- denoising
- motion field
- supervised learning
- predictive model
- anchovy catches
- support vector regression
- moving objects
- load forecasting
- restricted boltzmann machine
- neural network
- motion detection
- wind power
- human body
- feature points
- training set
- video sequences
- computer vision
- learning algorithm