arXiv subjects

Kazuya Hata

Publications and source records attributed to Kazuya Hata.

4 recordsLinked to original sources

arXiv2023

Retraining-free Customized ASR for Enharmonic Words Based on a Named-Entity-Aware Model and Phoneme Similarity Estimation

End-to-end automatic speech recognition (E2E-ASR) has the potential to improve performance, but a specific issue that needs to be addressed is the difficulty it has in handling enharmonic words: named entities (NEs) with the same pronunciation and part of speech that are spelled differently. This often occurs with Japanese personal names that have the same pronunciation but different Kanji characters. Since such NE words tend to be important keywords, ASR easily loses user trust if it misrecognizes them. To solve these problems, this paper proposes a novel retraining-free customized method for E2E-ASRs based on a named-entity-aware E2E-ASR model and phoneme similarity estimation. Experimental results show that the proposed method improves the target NE character error rate by 35.7% on average relative to the conventional E2E-ASR model when selecting personal names as a target NE.

cs.SD↗

arXiv2005

Hamiltonian of Homonucleus Molecules for NMR Quantum Computing

We derive the Hamiltonian in the rotating frame for NMR quantum computing with homonucleus molecules as its computational resource. The Hamiltonian thus obtained is different from conventional Hamiltonians that appear in literature. It is shown that control pulses designed for heteronucleus spins can be translated to pulses for homonucleus spins by simply replacing hard pulses by soft pulses with properly chosen pulse width. To demonstrate the validity of our Hamiltonian, we conducted several experiments employing cytosine as a homonucleus molecule. All the experimental results demonstrate that our Hamiltonian accurately describes the dynamics of the spins, while the conventional Hamiltonian fails. Finally we use our Hamiltonian for precise control of field inhomogeneity compensation with a pair of $π$-pulses.

quant-ph↗

arXiv2004

Warp-Drive Quantum Computation

Recently it has been shown that time-optimal quantum computation is attained by using the Cartan decomposition of a unitary matrix. We extend this approach by noting that the unitary group is compact. This allows us to reduce the execution time of a quantum algorithm $U_{\rm alg}$ further by adding an extra gate $W$ to it. This gate $W$ sends $U_{\rm alg}$ to another algorithm $WU_{\rm alg}$ which is executable in a shorter time than $U_{\rm alg}$. We call this technique warp-drive. Here we show both theoretically and experimentally that the execution time of Grover's algorithm is reduced in two-qubit NMR quantum computer. Warp-drive is potentially a powerful tool in accelerating algorithms and reducing the errors in any realization. of a quantum computer

quant-ph↗

arXiv2004

Demonstrating quantum algorithm acceleration with NMR quantum computer

In general, a quantum circuit is constructed with elementary gates, such as one-qubit gates and CNOT gates. It is possible, however, to speed up the execution time of a given circuit by merging those elementary gates together into larger modules, such that the desired unitary matrix expressing the algorithm is directly implemented. We demonstrate this by taking the two-qubit Grover's algorithm implemented in NMR quantum computation, whose pseudopure state is generated by cyclic permutations of the state populations. This is the first exact time-optimal solution, to our knowledge, obtained for a self-contained quantum algorithm.

quant-ph↗