Automatic segmentation and labeling of continuous speech without bootstrapping | IEEE Conference Publication | IEEE Xplore