In this paper we describe a method to perform sequencediscriminative training of neural network acoustic models without the need for frame-level cross-entropy pre-training. We use the lattice-free version of the maximum mutual information (MMI) criterion: LF-MMI. To make its computation feasible we use a phone n-gram language model, in place of the word language model. To further reduce its space and time complexity we compute the objective function using neural network outputs at one third the standard frame rate. These changes enable us to perform the computation for the forward-backward algorithm on GPUs. Further the reduced output frame-rate also provides a significant speed-up during decoding. We present results on 5 different LVCSR tasks with training data ranging from 100 to 2100 hours. Models trained with LF-MMI provide a relative word error rate reduction of ∼11.5%, over those trained with cross-entropy objective function, and ∼8%, over those trained with cross-entropy and sMBR objective functions. A further reduction of ∼2.5%, relative, can be obtained by fine tuning these models with the word-lattice based sMBR objective function.
We describe in this paper the experiences of the Johns Hopkins University team during the inaugural DIHARD diarization evaluation. This new task provided microphone recordings in a variety of difficult conditions and challenged researchers to fully consider all speaker activity, without the currently typical practices of unscored collars or ignored overlapping speaker segments. This paper explores several key aspects of currently state-of-the-art diarization methods, such as training data selection, signal bandwidth for feature extraction, representations of speech segments (i-vector versus x-vector), and domainadaptive processing. In the end, our best system clustered xvector embeddings trained on wideband microphone data followed by Variational-Bayesian refinement, and a speech activity detector specifically trained for this task with in-domain data was found to be the best performing. After presenting these decisions and their final result, we discuss lessons learned and remaining challenges within the lens of this new approach to diarization performance measurement.
Long Short-Term Memory networks (LSTMs) are a component of many state-of-the-art DNN-based speech recognition systems. Dropout is a popular method to improve generalization in DNN training. In this paper we describe extensive experiments in which we investigated the best way to combine dropout with LSTMs-specifically, projected LSTMs (LSTMP). We investigated various locations in the LSTM to place the dropout (and various combinations of locations), and a variety of dropout schedules. Our optimized recipe gives consistent improvements in WER across a range of datasets, including Switchboard, TED-LIUM and AMI.Projected LSTMs (LSTMPs) [4] are an important component of our baseline system, and to provide context for our explanation of dropout we will repeat the equations for them; here xt is the
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.