Forecast-MAE: Self-supervised Pre-training for Motion Forecasting with Masked Autoencoders.

Jie Cheng, Xiaodong Mei, Ming Liu

Browse the full ICCV paper archive.