A low latency ASR-free end to end spoken language understanding system

Mhiri, Mohamed; Myer, Samuel; Tomar, Vikrant Singh

doi:10.48550/arxiv.2011.04884

Cited by 1 publication

(2 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…RNN+Pre-training [9] 98.80 CNN+Segment pooling [1] 97.80 CNN+GRU(SotA) [21] 99.10 3D-CNN+LSTM+CE 99.26 kernels are used in the first layer, followed by 32 in the second layer. As depicted in Figure 3, the temporal dynamics are preserved.…”

Section: Model Intentmentioning

confidence: 99%

See 1 more Smart Citation

Sequential End-to-End Intent and Slot Label Classification and Localization

Cao

Potdar

Avila

2021

Preprint

View full text Add to dashboard Cite

Human-computer interaction (HCI) is significantly impacted by delayed responses from a spoken dialogue system. Hence, endto-end (e2e) spoken language understanding (SLU) solutions have recently been proposed to decrease latency. Such approaches allow for the extraction of semantic information directly from the speech signal, thus bypassing the need for a transcript from an automatic speech recognition (ASR) system. In this paper, we propose a compact e2e SLU architecture for streaming scenarios, where chunks of the speech signal are processed continuously to predict intent and slot values. Our model is based on a 3D convolutional neural network (3D-CNN) and a unidirectional long short-term memory (LSTM). We compare the performance of two alignment-free losses: the connectionist temporal classification (CTC) method and its adapted version, namely connectionist temporal localization (CTL). The latter performs not only the classification but also localization of sequential audio events. The proposed solution is evaluated on the Fluent Speech Command dataset and results show our model ability to process incoming speech signal, reaching accuracy as high as 98.97 % for CTC and 98.78 % for CTL on single-label classification, and as high as 95.69 % for CTC and 95.28 % for CTL on two-label prediction.

show abstract

Section: Model Intentmentioning

confidence: 99%

“…Spoken language understanding (SLU) aims at extracting structured semantic representations, such as intent and slots, from the speech signal [1]. These representations are crucial to enable speech as the primary mode of human-computer interaction (HCI) [2].…”

Section: Introductionmentioning

confidence: 99%