Syllable language models for Mandarin speech recognition: exploiting character language models.

Mandarin Chinese is based on characters which are syllabic in nature and morphological in meaning. All spoken languages have syllabiotactic rules which govern the construction of syllables and their allowed sequences. These constraints are not as restrictive as those learned from word sequences, but they can provide additional useful linguistic information. Hence, it is possible to improve speech recognition performance by appropriately combining these two types of constraints. For the Chinese language considered in this paper, character level language models (LMs) can be used as a first level approximation to allowed syllable sequences. To test this idea, word and character level n-gram LMs were trained on 2.8 billion words (equivalent to 4.3 billion characters) of texts from a wide collection of text sources. Both hypothesis and model based combination techniques were investigated to combine word and character level LMs. Significant character error rate reductions up to 7.3% relative were obtained on a state-of-the-art Mandarin Chinese broadcast audio recognition task using an adapted history dependent multi-level LM that performs a log-linearly combination of character and word level LMs. This supports the hypothesis that character or syllable sequence models are useful for improving Mandarin speech recognition performance.

Energy recycling from multigigahertz clocks using fully integrated switching converters

Large digital chips use a significant amount of energy to broadcast a low-skew, multigigahertz clock to millions of latches located throughout the chip. Every clock cycle, the large aggregate capacitance of the clock network is charged from the supply and then discharged to ground. Instead of wasting this stored energy, it is possible to recycle the energy by controlling its delivery to another part of the chip using an on-chip dc-dc converter. The clock driver and switching converter circuits share many compatible characteristics that allow them to be merged into a single design and fully integrated on-chip. Our buck converter prototype, manufactured in 90-nm CMOS, provides a proof-of-concept that clock network energy can be recycled to other parts of the chip, thus lowering overall energy consumption. It also confirms that monolithic multigigahertz switching converters utilizing zero-voltage switching can be implemented in deep-submicrometer CMOS. With multigigahertz operation, fully integrated inductors and capacitors use a small amount of chip area with low losses. Combining the clock driver with the power converter can share the large MOSFET drivers necessary as well as being energy and space efficient. We present an analysis of the losses which we confirm by experimentally comparing the merged circuit with a conventional clock driver. © 2012 IEEE.

Paraphrastic language models and combination with neural network language models

In natural languages multiple word sequences can represent the same underlying meaning. Only modelling the observed surface word sequence can result in poor context coverage, for example, when using n-gram language models (LM). To handle this issue, paraphrastic LMs were proposed in previous research and successfully applied to a US English conversational telephone speech transcription task. In order to exploit the complementary characteristics of paraphrastic LMs and neural network LMs (NNLM), the combination between the two is investigated in this paper. To investigate paraphrastic LMs' generalization ability to other languages, experiments are conducted on a Mandarin Chinese broadcast speech transcription task. Using a paraphrastic multi-level LM modelling both word and phrase sequences, significant error rate reductions of 0.9% absolute (9% relative) and 0.5% absolute (5% relative) were obtained over the baseline n-gram and NNLM systems respectively, after a combination with word and phrase level NNLMs. © 2013 IEEE.