Search

This paper introduces a spectral analysis of time-seires data derived from real-time time-dependent density functional theory (TDDFT) using Singular Spectrum Analysis (SSA). TDDFT is a robust method for obtaining molecular excited states and optical spectra by tracking the time evolution of dynamical dipole moments. However, the spectral resolution can be compromised when Fourier transformation's total time duration is insufficient. SSA enabled the extraction of specific oscillation components from the time-series data, facilitating the generation of higher-precision spectra. Even with relatively short time-series dataset, the predictive extension of SSA yielded high-resolution spectra, demonstrating substantial agreement with results obtained through conventional methods. The efficacy of this approach was validated for several small molecules, including ethylene, benzene, and others. SSA's ability to conduct detailed spectral anasysis in specific energy regions enhance spectral resolution and facilitates the clarification of oscillation components within these regions. Real-time TDDFT combined with SSA provides a new analytical method for analyzing the optical properties of molecules, significantly improving the accuracy of the analysis of emission and absorption spectra analysis. This method is expected to have various applications., Comment: 11 pages 5 png figures

Resonance fluorescence spectra of a driven Kerr nonlinear resonator is investigated both theoretically and experimentally. When the Kerr nonlinear resonator is driven strongly such that the induced Rabi frequency is comparable to or larger than the Kerr nonlinearity, the system cannot be approximated as a two-level system. We theoretically derive characteristic features in the fluorescence spectra such as the decrease of the center-peak intensity and the asymmetric sideband peaks in the presence of finite dephasing. Those features are consistently explained by the population of the initial dressed state and its transition matrix element to the final dressed state of the transition corresponding to each peak. Finally, we experimentally measure the resonance fluorescence spectra of a driven superconducting Kerr nonlinear resonator and find a quantitative agreement with our theory., Comment: 18 pages, 10 figures

Background: Recent advances in large language models highlight the need for high-quality multilingual medical datasets. While Japan leads globally in CT scanner deployment and utilization, the lack of large-scale Japanese radiology datasets has hindered the development of specialized language models for medical imaging analysis. Objective: To develop a comprehensive Japanese CT report dataset through machine translation and establish a specialized language model for structured finding classification. Additionally, to create a rigorously validated evaluation dataset through expert radiologist review. Methods: We translated the CT-RATE dataset (24,283 CT reports from 21,304 patients) into Japanese using GPT-4o mini. The training dataset consisted of 22,778 machine-translated reports, while the validation dataset included 150 radiologist-revised reports. We developed CT-BERT-JPN based on "tohoku-nlp/bert-base-japanese-v3" architecture for extracting 18 structured findings from Japanese radiology reports. Results: Translation metrics showed strong performance with BLEU scores of 0.731 and 0.690, and ROUGE scores ranging from 0.770 to 0.876 for Findings and from 0.748 to 0.857 for Impression sections. CT-BERT-JPN demonstrated superior performance compared to GPT-4o in 11 out of 18 conditions, including lymphadenopathy (+14.2%), interlobular septal thickening (+10.9%), and atelectasis (+7.4%). The model maintained F1 scores exceeding 0.95 in 14 out of 18 conditions and achieved perfect scores in four conditions. Conclusions: Our study establishes a robust Japanese CT report dataset and demonstrates the effectiveness of a specialized language model for structured finding classification. The hybrid approach of machine translation and expert validation enables the creation of large-scale medical datasets while maintaining high quality., Comment: Dataset available at https://huggingface.co/datasets/YYama0/CT-RATE-JPN

Robot controllers are often optimised for a single robot in a single environment. This approach proves brittle, as such a controller will often fail to produce sensible behavior for a new morphology or environment. In comparison, animal gaits are robust and versatile. By observing animals, and attempting to extract general principles of locomotion from their movement, we aim to design a single decentralised controller applicable to diverse morphologies and environments. The controller implements the three components 1) undulation, 2) peristalsis, and 3) leg motion, which we believe are the essential elements in most animal gaits. The controller is tested on a variety of simulated centipede-like robots. The centipede is chosen as inspiration because it moves using both body contractions and legged locomotion. For a controller to work in qualitatively different settings, it must also be able to exhibit qualitatively different behaviors. We find that six different modes of locomotion emerge from our controller in response to environmental and morphological changes. We also find that different parts of the centipede model can exhibit different modes of locomotion, simultaneously, based on local morphological features. This controller can potentially aid in the design or evolution of robots, by quickly testing the potential of a morphology, or be used to get insights about underlying locomotion principles in the centipede.

JASMINE is a Japanese planned space mission that aims to reveal the formation history of our Galaxy and discover habitable exoEarths. For these objectives, the JASMINE satellite performs high-precision astrometric observations of the Galactic bulge and high-precision transit monitoring of M-dwarfs in the near-infrared (1.0-1.6 microns in wavelength). For feasibility studies, we develop an image simulation software named JASMINE-imagesim, which produces realistic observation images. This software takes into account various factors such as the optical point spread function (PSF), telescope jitter caused by the satellite's attitude control error (ACE), detector flat patterns, exposure timing differences between detector pixels, and various noise factors. As an example, we report a simulation for the feasibility study of astrometric observations using JASMINE-imagesim. The simulation confirms that the required position measurement accuracy of 4 mas for a single exposure of 12.5-mag objects is achievable if the telescope pointing jitter uniformly dilutes the PSF across all stars in the field of view. On the other hand, the simulation also demonstrates that the combination of realistic pointing jitter and exposure timing differences in the detector can significantly degrade accuracy and prevent achieving the requirement. This means that certain countermeasures against this issue must be developed. This result implies that this kind of simulation is important for mission planning and advanced developments to realize more realistic simulations help us to identify critical issues and also devise effective solutions., Comment: 13 pages, 7 figures, 1 table

This paper introduces a novel framework for Archetypal Analysis (AA) tailored to ordinal data, particularly from questionnaires. Unlike existing methods, the proposed method, Ordinal Archetypal Analysis (OAA), bypasses the two-step process of transforming ordinal data into continuous scales and operates directly on the ordinal data. We extend traditional AA methods to handle the subjective nature of questionnaire-based data, acknowledging individual differences in scale perception. We introduce the Response Bias Ordinal Archetypal Analysis (RBOAA), which learns individualized scales for each subject during optimization. The effectiveness of these methods is demonstrated on synthetic data and the European Social Survey dataset, highlighting their potential to provide deeper insights into human behavior and perception. The study underscores the importance of considering response bias in cross-national research and offers a principled approach to analyzing ordinal data through Archetypal Analysis., Comment: Accepted at Machine Learning and Signal Processing 2024

We present a distant automatic speech recognition (DASR) system developed for the CHiME-8 DASR track. It consists of a diarization first pipeline. For diarization, we use end-to-end diarization with vector clustering (EEND-VC) followed by target speaker voice activity detection (TS-VAD) refinement. To deal with various numbers of speakers, we developed a new multi-channel speaker counting approach. We then apply guided source separation (GSS) with several improvements to the baseline system. Finally, we perform ASR using a combination of systems built from strong pre-trained models. Our proposed system achieves a macro tcpWER of 21.3 % on the dev set, which is a 57 % relative improvement over the baseline., Comment: 5 pages, 4 figures, CHiME8 challenge

I describe in this paper how the shallow water waves of Airy, produced by submarine earthquake, gentle on a deep open sea traveling by the cruising speed, get instantaneous $ +\infty$ or $-\infty$ traveling speed in approaching the shallow water near the coasts at the points of the surface on x=X, near the crest or the trough, where the tangent of itself and the tangent of the seabed coincide, rushing as inland tunamis ( near the crest) or as offshore tunamis (near the trough).

This paper introduces a novel approach called sentence-wise speech summarization (Sen-SSum), which generates text summaries from a spoken document in a sentence-by-sentence manner. Sen-SSum combines the real-time processing of automatic speech recognition (ASR) with the conciseness of speech summarization. To explore this approach, we present two datasets for Sen-SSum: Mega-SSum and CSJ-SSum. Using these datasets, our study evaluates two types of Transformer-based models: 1) cascade models that combine ASR and strong text summarization models, and 2) end-to-end (E2E) models that directly convert speech into a text summary. While E2E models are appealing to develop compute-efficient models, they perform worse than cascade models. Therefore, we propose knowledge distillation for E2E models using pseudo-summaries generated by the cascade models. Our experiments show that this proposed knowledge distillation effectively improves the performance of the E2E model on both datasets., Comment: Accepted to Interspeech2024. Dataset: https://huggingface.co/datasets/komats/mega-ssum

This paper describes NAIST's submission to the simultaneous track of the IWSLT 2024 Evaluation Campaign: English-to-{German, Japanese, Chinese} speech-to-text translation and English-to-Japanese speech-to-speech translation. We develop a multilingual end-to-end speech-to-text translation model combining two pre-trained language models, HuBERT and mBART. We trained this model with two decoding policies, Local Agreement (LA) and AlignAtt. The submitted models employ the LA policy because it outperformed the AlignAtt policy in previous models. Our speech-to-speech translation method is a cascade of the above speech-to-text model and an incremental text-to-speech (TTS) module that incorporates a phoneme estimation model, a parallel acoustic model, and a parallel WaveGAN vocoder. We improved our incremental TTS by applying the Transformer architecture with the AlignAtt policy for the estimation model. The results show that our upgraded TTS module contributed to improving the system performance., Comment: IWSLT 2024 system paper

Large language models (LLMs) have been successfully applied for rescoring automatic speech recognition (ASR) hypotheses. However, their ability to rescore ASR hypotheses of casual conversations has not been sufficiently explored. In this study, we reveal it by performing N-best ASR hypotheses rescoring using Llama2 on the CHiME-7 distant ASR (DASR) task. Llama2 is one of the most representative LLMs, and the CHiME-7 DASR task provides datasets of casual conversations between multiple participants. We investigate the effects of domain adaptation of the LLM and context carry-over when performing N-best rescoring. Experimental results show that, even without domain adaptation, Llama2 outperforms a standard-size domain-adapted Transformer-LM, especially when using a long context. Domain adaptation shortens the context length needed with Llama2 to achieve its best performance, i.e., it reduces the computational cost of Llama2., Comment: 5 pages

The extraction of variable definitions from scientific and technical papers is essential for understanding these documents. However, the characteristics of variable definitions, such as the length and the words that make up the definition, differ among fields, which leads to differences in the performance of existing extraction methods across fields. Although preparing training data specific to each field can improve the performance of the methods, it is costly to create high-quality training data. To address this challenge, this study proposes a new method that generates new definition sentences from template sentences and variable-definition pairs in the training data. The proposed method has been tested on papers about chemical processes, and the results show that the model trained with the definition sentences generated by the proposed method achieved a higher accuracy of 89.6%, surpassing existing models.

Aiming to simulate the radiation damage effect on a dual alpha+beta phase Ti-6Al-4V alloy utilized as high-intensity accelerator beam window material, a series of irradiation experiments were conducted with a 2.8 MeV-Fe^2+ ion beam in several dpa regions at room temperature. The nano-indentation hardness increased steeply at 1 dpa and unchanged up to 10 dpa, due to the saturation of defect clusters and tangled dislocations in the dominant alpha-phase matrix with a size of 2~3 nm and a density of about 1x10^23 m^-3. In contrast in the intergranular beta-phase, larger loops of 20~30 nm diameter were observed with much less density of about 5x10^20 m^-3. The diffraction pattern showed rectilinear diffuse streaks between the beta-phase reflections, corresponding to the omega-phase precursor, without dose dependency in its intensity. FFT/I-FFT analysis of the HREM revealed a sub-nanometer-sized lattice disorder with local fluctuations, not discrete but continuous, and homogeneously distributed within the matrix beta-phase stably against the irradiation. The significantly low dislocation density and the absence of phase transformation in the beta-phase matrix could be attributed either to the strong sink effect expected for this distinctive sub-nanometer-sized homogeneous lattice disorder or to the anomalous point defect recombination induced by the high mobility of vacancies, both of which are originated from the metastable omega-phase precursors specifically formed in the beta(BCC) phase of group-4 transition metals., Comment: 36 pages, 15 figures

We continue to study the unbounded $\mathfrak{sl}_3$-laminations [IK22], with a focus on their structures at punctures. A key ingredient is their relation to the root data of $\mathfrak{sl}_3$. After giving a classification of signed $\mathfrak{sl}_3$-webs around a puncture, we describe the tropicalization of the Goncharov--Shen's Weyl group action in detail. We also clarify the relationship with several other approaches by Shen--Sun--Weng [SSW23] and Fraser--Pylyavskyy [FP21]. Finally, we discuss a formulation of unbounded $\mathfrak{g}$-laminations for a general semisimple Lie algebra $\mathfrak{g}$ in brief., Comment: 57 pages, 26 figures. v2: added a comment on the Roger--Yang relations in p.42. arXiv admin note: text overlap with arXiv:2204.08947

The cluster algebra associated with an acyclic quiver has a special mutation loop $\tau$, called the cluster Donaldson--Thomas (DT) transformation, related to the Auslander--Reiten translation. In this paper, we characterize the finite-tame-wild trichotomy for acyclic quivers by the sign stability of $\tau$ introduced in [IK21] and its cluster stretch factor. As an application, we compute several kinds of entropies of $\tau$ and other mutation loops. In particular, we show that the algebraic and categorical entropies of $\tau$ are commonly given by the logarithm of the spectral radius of the Coxeter matrix associated with the quiver, and that any mutation loop of finite or tame acyclic quivers have zero algebraic entropy., Comment: 23 pages, 3 figures

Confidence estimation, in which we estimate the reliability of each recognized token (e.g., word, sub-word, and character) in automatic speech recognition (ASR) hypotheses and detect incorrectly recognized tokens, is an important function for developing ASR applications. In this study, we perform confidence estimation for end-to-end (E2E) ASR hypotheses. Recent E2E ASR systems show high performance (e.g., around 5% token error rates) for various ASR tasks. In such situations, confidence estimation becomes difficult since we need to detect infrequent incorrect tokens from mostly correct token sequences. To tackle this imbalanced dataset problem, we employ a bidirectional long short-term memory (BLSTM)-based model as a strong binary-class (correct/incorrect) sequence labeler that is trained with a class balancing objective. We experimentally confirmed that, by utilizing several types of ASR decoding scores as its auxiliary features, the model steadily shows high confidence estimation performance under highly imbalanced settings. We also confirmed that the BLSTM-based model outperforms Transformer-based confidence estimation models, which greatly underestimate incorrect tokens., Comment: Accepted to ICASSP 2021

We propose a skein model for the quantum cluster algebras of surface type with coefficients. We introduce a skein algebra $\mathscr{S}_{\Sigma,\mathbb{W}}^{A}$ of a walled surface $(\Sigma,\mathbb{W})$, and prove that it has a quantum cluster structure. The walled surfaces naturally generalize the marked surfaces with multi-laminations, which have been used to describe the quantum cluster algebras of geometric type for marked surfaces by Fomin--Thurston [FT18]. Moreover, we give skein theoretic interpretation for some of quasi-homomorphisms [Fra16] between these quantum cluster algebras., Comment: 43 pages; v2: Major revision, introduction refreshed, arguments concerning the minimal position moved to Section 5, the Fomin--Thurston's description of the normalized cluster algebras moved to Section 4.2, Examples 2.13 and 2.14 added

Simultaneous translation is a task in which the translation begins before the end of an input speech segment. Its evaluation should be conducted based on latency in addition to quality, and for users, the smallest possible amount of latency is preferable. Most existing metrics measure latency based on the start timings of partial translations and ignore their duration. This means such metrics do not penalize the latency caused by long translation output, which delays the comprehension of users and subsequent translations. In this work, we propose a novel latency evaluation metric for simultaneous translation called \emph{Average Token Delay} (ATD) that focuses on the duration of partial translations. We demonstrate its effectiveness through analyses simulating user-side latency based on Ear-Voice Span (EVS). In our experiment, ATD had the highest correlation with EVS among baseline latency metrics under most conditions., Comment: Extended version of the paper (doi: 10.21437/Interspeech.2023-933) which appeared in INTERSPEECH 2023