Motion-Corrected Moving Average (MCMA) improves video segmentation by incorporating temporal information without altering the model architecture.
