DOMA: Neural Motion Prediction for RF Targets

RF Quantum SCYTHE lightweight DOMA (Dynamic Object
Motion Analysis) model for short-horizon motion prediction of
RF targets from kinematic traces. Our pipeline trains a compact
MLP on synthetic trajectories and evaluates with autoregressive
rollouts to report Average Displacement Error (ADE) and Final
Displacement Error (FDE). All figures and tables are autogenerated for reproducibility. DOMA achieves competitive performance against Kalman filter baselines while maintaining subsecond training times.