RNN based Incremental Online Spoken Language Understanding
This work addresses latency issues in SLU systems for real-time spoken interactions, but it is incremental as it builds on existing RNN methods with a focus on online processing.
The paper tackles the high latency in Spoken Language Understanding (SLU) caused by sequential processing of ASR and NLU modules by proposing RNN-based incremental processing for intent detection, achieving lower latencies without significant accuracy reduction on benchmark datasets like ATIS, Snips, and Facebook's multilingual dialog datasets.
Spoken Language Understanding (SLU) typically comprises of an automatic speech recognition (ASR) followed by a natural language understanding (NLU) module. The two modules process signals in a blocking sequential fashion, i.e., the NLU often has to wait for the ASR to finish processing on an utterance basis, potentially leading to high latencies that render the spoken interaction less natural. In this paper, we propose recurrent neural network (RNN) based incremental processing towards the SLU task of intent detection. The proposed methodology offers lower latencies than a typical SLU system, without any significant reduction in system accuracy. We introduce and analyze different recurrent neural network architectures for incremental and online processing of the ASR transcripts and compare it to the existing offline systems. A lexical End-of-Sentence (EOS) detector is proposed for segmenting the stream of transcript into sentences for intent classification. Intent detection experiments are conducted on benchmark ATIS, Snips and Facebook's multilingual task oriented dialog datasets modified to emulate a continuous incremental stream of words with no utterance demarcation. We also analyze the prospects of early intent detection, before EOS, with our proposed system.