Fine-Tuning of Pre-Trained End-to-End Speech Recognition with Generative Adversarial Networks | IEEE Conference Publication | IEEE Xplore