The Augmented Extended Kalman Filter (AEKF) enables simultaneous estimation of structural responses and system parameters but faces significant computational challenges for large structural systems. This paper presents an efficient AEKF implementation combining adaptive Jacobian calculation with spectral Kalman filter transition for optimized computation. The adaptive strategy selectively updates the Jacobian matrix based on parameter convergence statistics, while the spectral transition reduces computational demands during stable estimation phases. Validation using a three-span bridge (408 DOF) and a 20-story frame structure (378 DOF) under seismic excitation demonstrates that the adaptive strategy reduces processing time by 52% while maintaining displacement estimation accuracy within 5%, with spectral transition providing an additional 6% reduction during stable phases. For the bridge model, overall assessment time decreases from 146 to 70 min while preserving estimation accuracy. Similar improvements are achieved for the frame structure, with computation time reduced from 138 to 65 min. These improvements enable more practical implementation of AEKF-based structural response estimation for large-scale systems while preserving assessment accuracy.