This paper presents a framework of using Chinese speech to access images via English captions. The formulation and the structure mapping rules of Chinese and English named entities are extracted from an NICT foreign location name corpus. For a named location, name part and keyword part are usually transliterated and translated, respectively. Keyword spotting identifies the keyword from speech q...

Compared to English, Chinese has a simpler and more restricted syllabic structure. In order to exploit the special characteristics of Chinese, syllable is selected as the unit for ASR lattice representation. For the sake of fast retrieval, syllable lattices are clustered into confusion network linear lattices, and then encoded into inverted index. To recover the posterior probabilities of prune...

Generally, exact matching is widely used for keyword spotting (KWS). Its performance depends heavily on the recognition accuracy. As for phone-based KWS system, the influence of phoneme error rate (PER) on KWS increases as the length of phoneme sequence for the keyword grows. Approximate matching is an alteration to compensate errors in recognition. Compared to exact matching, the calculation c...

The intrinsic advantages of whole-word acoustic modeling are offset by the problem of data sparsity. To address this, we present several parametric approaches to estimating intra-word phonetic timing models under the assumption that relative timing is independent of word duration. We show evidence that the timing of phonetic events is well described by the Gaussian distribution. We explore the ...

We submitted a system composed of 26 subsystems as the required run. 13 subsystems are based on Acoustic Keyword Spotting and 13 on DTW. All of them were using three state phoneme posteriors as input. The underlaying phoneme posterior estimators were both in-language (Czech, English) and out-of-language (other 12 languages). We also performed unsupervised adaptation of the artificial neural net...

To spot keywords on handwritten documents, we present a hybrid keyword spotting system, based on features extracted with Convolutional Deep Belief Networks and using Dynamic Time Warping for word scoring. Features are learned from word images, in an unsupervised manner, using a sliding window to extract horizontal patches. For two single writer historical data sets, it is shown that the propose...

A hierarchical framework is proposed to address the issues of modeling different type of words in keyword spotting (KWS). Keyword models are built at various levels according to the availability of training set resources for each individual word. The proposed approach improves the performance of KWS even when no training speech is available for the keywords. It also suggests an easier way to co...

\ , I \ I I The problem Of discriminating keyword and non-keyword speech which is important in wordspotting applications is addressed here. We have shown that garbage models cannot reduce both rejection and false alarm rates simultaneously. Thus, the following relation becomes apparent from the above inequalities O'In I ) > " ( o ' l l P ) . P(08 I R P ) > p(oP I R 1 ) (4) TO achieve this we ha...

This paper describes an expansion of the Viterbi algorithm in a cursive script word recognizer based on Hidden Markov Models (HMM) that allows the detection of subwords. The advantage of word recognizers based on HMMs is that there is no need for a segmentation of the word image into single letters. In common systems segmentation still has to take place on a word level which rises problems in m...

This paper describes several ways of keywords spotting (KWS), based on Gaussian mixture (GM) hidden Markov modelling (HMM). Context-independent and dependent phoneme models are used in our system. The system was trained and evaluated on informal continuous speech. We used different complexities of KWS recognition networks and different types of phoneme models. The impact of these parameters on ...