SPformer: Hybrid Sequential-Parallel Architectures for Automatic Speech Recognition | IEEE Conference Publication | IEEE Xplore