1: \begin{abstract}

2: 		% 140 words

3: 		We present a novel end-to-end deep learning-based adaptation control algorithm for frequency-domain adaptive system identification. The proposed method exploits a deep neural network to map observed signal features to corresponding step-sizes which control the filter adaptation. The parameters of the network are optimized in an end-to-end fashion by minimizing the average \commentTHd{normalized} system distance of the adaptive filter. This avoids the need of explicit signal power spectral density estimation as required for model-based adaptation control and further auxiliary mechanisms to deal with model inaccuracies.

4: 		The proposed algorithm achieves fast convergence and robust steady-state performance for scenarios characterized by \commentTHf{high-level,} non-white and non-stationary \commentTHf{additive} noise signals, \commentTHf{abrupt} environment changes and additional model inaccuracies. 

5: 	\end{abstract}

6: