In keyword spotting from handwritten documents by text query, the word similarity is usually computed by combining character similarities, which are desired to approximate the logarithm of the character probabilities. In this paper, we propose to directly estimate the posterior probability (also called confidence) of candidate characters based on the N-best paths from the candidate segmentation...

The automatic transcription of unconstrained continuous handwritten text requires well trained recognition systems. The semi-supervised paradigm introduces the concept of not only using labeled data but also unlabeled data in the learning process. Unlabeled data can be gathered at little or not cost. Hence it has the potential to reduce the need for labeling training data, a tedious and costly ...

In this paper, we propose a method for text-query-based keyword spotting from online Chinese handwritten documents using character classi ̄cation model. The similarity between the query word and handwriting is obtained by combining the character classi ̄cation scores. The classi ̄er is trained by one-versus-all strategy so that it gives high similarity to the target class and low scores to the oth...

Spoken keyword spotting has been widely used to facilitate an always-on voice interface in consumer electronics, owing its simplicity and low latency. Small-footprint based on tiny convolutional neural networks can be implemented resource-constrained but energy-efficient microcontrollers real time. However, it is difficult for learn the noise-robustness properties essential successful interface...

We present the preliminary results of applying a set of parameters of the AM-FM model for recognizing word utterances. By acquiring modulation based parameters from the amplitude envelope (AE) and the instantaneous frequency – both obtained by demodulating at four selected center frequencies – a compact feature set is created for each frame of a word utterance. Applying a dynamic time warping o...

This paper proposes several improvements to multilingual training of neural network acoustic models for speech recognition and keyword spotting in the context of low-resource languages. We concentrate on the stacked architecture where the first network is used as a bottleneck feature extractor and the second network as the acoustic model. We propose to improve multilingual training when the amo...

In this paper, we propose an attention-based end-to-end neural approach for small-footprint keyword spotting (KWS), which aims to simplify the pipelines of building a production-quality KWS system. Our model consists of an encoder and an attention mechanism. The encoder transforms the input signal into a high level representation using RNNs. Then the attention mechanism weights the encoder feat...