13197 research outputs found
Sort by
Absolute evaluation of internal and external quantum efficiencies and light extraction efficiency in InGaN single quantum wells by simultaneous photoacoustic and photoluminescence measurements combined with integrating-sphere method
No full textSeparated evaluation of factors in the external quantum efficiency (EQE) is important in order to improve the characteristics of semiconductors optical devices. Especially, the internal quantum efficiency (IQE) is an important value which indicates crystal quality of the active layers, and an accurate method for estimating the IQE values is required. The IQE is usually estimated from temperature dependence of photoluminescence (PL) intensity by assuming that the IQE at cryogenic temperature is 100%. However, III-nitride semiconductor materials, used in many optical devices, usually have large defect density, and the assumption is not necessarily valid. In our previous report, we demonstrated the simultaneous photoacoustic (PA) and PL measurements to accurately estimate the IQE values in GaN films with various qualities and obtained reasonable results. In this work, we have successfully realized reproducible measurements with high accuracy for an InGaN-QW sample by suppressing the background noise significantly. Furthermore, we have also measured the values of EQE by using an integrating-sphere. Since the light extraction efficiency (LEE) can be obtained by the values of IQE and EQE, it has been shown that the overall picture of emission efficiency can be provided by our method.journal articl
Chemical Graph-Based Transformer Models for Yield Prediction of High-Throughput Cross-Coupling Reaction Datasets
No full textThe chemical reaction yield is an important factor to determine the reaction conditions. Recently, many data-driven models for yield prediction using high-throughput experimentation datasets have been reported. In this study, we propose a neural network architecture based on the chemical graphs of the reaction components to predict the reaction yield. The proposed model is the sequential combination of a message-passing neural network and a transformer encoder (MPNN-Transformer). The reaction components are converted to molecular matrices by the first network, followed by the interplay of the reaction components in the second network after adding the embeddings of the compound roles in the chemical reaction. The predictive ability of the proposed models was compared with state-of-the-art yield prediction models using two high-throughput experimental datasets: the Buchwald2013Miyaura cross-coupling (SMC) reaction datasets. Overall, the MPNN-Transformer models showed high prediction accuracy for the BHC reaction datasets and some of the extrapolation-oriented SMC reaction datasets. These models also performed well when the training dataset size was relatively large. Furthermore, analyzing the poorly predicted reactions for the BHC reaction dataset revealed a limitation of the data-driven yield prediction approach based on the chemical structural similarity.journal articl
Fission yeast Pib2 localizes to vacuolar membranes and regulates TOR complex 1 through evolutionarily conserved domains
Get PDFTOR complex 1 (TORC1) is a multi-protein kinase complex that coordinates cellular growth with environmental cues. Recent studies have identified Pib2 as a critical activator of TORC1 in budding yeast. Here, we show that loss of Pib2 causes severe growth defects in fission yeast cells, particularly when basal TORC1 activity is diminished by hypomorphic mutations in tor2, the gene encoding the catalytic subunit of TORC1. Consistently, TORC1 activity is significantly compromised in the tor2 hypomorphic mutants lacking Pib2. Moreover, as in budding yeast, fission yeast Pib2 localizes to vacuolar membranes via its FYVE domain, with its tail motif indispensable for TORC1 activation. These results strongly suggest that Pib2-mediated positive regulation of TORC1 is evolutionarily conserved between the two yeast species.journal articl
Particle behaviors of Imipenem/Cilastatin in various liquids investigated by chemical structures and temporal microscope observations
No full textImipenem/Cilastatin (IPM/CS) is currently used as an embolic agent in transarterial embolization for chronic musculoskeletal pain due to its efficacy of short-term blood flow occlusion. However, due to its off-label use, the development of new materials to substitute IPM/CS is needed. Therefore, investigating the behavior of IPM/CS particles becomes crucial. In this paper, we observed the changes in particle size of IPM/CS when dissolved in phosphate-buffered saline, normal saline, iodinated contrast agent and water. The results indicated that the particle size of IPM/CS ranged from 16 to 5700B5m and has fastest hydrolysis rate in iodinated contrast agent. And we analyzed the possible structure of hydrolyzed product of IPM/CS in water at room temperature and 37 ℃ using 1HNMR, HPLC. The results suggested Cilastatin is stable at both temperatures, while Imipenem hydrolyzed to produce slightly different products at different temperatures.journal articl
Bidirectional Transformer Reranker for Grammatical Error Correction
No full textPre-trained sequence-to-sequence (seq2seq) models have achieved state-of-the-art results in the grammatical error correction tasks. However, these models are plagued by prediction bias owing to their unidirectional decoding. Thus, this study proposed a bidirectional transformer reranker (BTR) that re-estimates the probability of each candidate sentence generated by the pre-trained seq2seq model. The BTR preserves the seq2seq-style transformer architecture but utilizes a BERT-style self-attention mechanism in the decoder to compute the probability of each target token using masked language modeling to capture bidirectional representations from the target context. To guide the reranking process, the BTR adopted negative sampling in the objective function to minimize the unlikelihood. During inference, the BTR yielded the final results after comparing the reranked top-1 results with the original ones using an acceptance threshold λ. Experimental results showed that, when reranking candidates from a pre-trained seq2seq model, the T5-base, the BTR on top of T5-base yielded scores of 65.47 and 71.27 F0.5 on the CoNLL-14 and building educational applications 2019 (BEA) test sets, respectively, and yielded 59.52 GLEU score on the JFLEG corpus, with improvements of 0.36, 0.76, and 0.48 points compared with the original T5-base. Furthermore, when reranking candidates from T5-large, the BTR on top of T5-base improved the original T5-large by 0.26 on the BEA test set.journal articl
Emotional studies in dogs and cats and their estimation techniques: an engineering perspective
No full textDogs and cats have exceptionally developed sensory systems and abilities to recognize human sig-nals and emotional states. It makes them invaluable in roles such as working dogs and therapyanimals in human society. Understanding each other’s emotional state is essential to working withthem effectively. However, the low accuracy of human emotional estimation in dogs and cats isa significant issue. Due to individual differences and cognitive biases affecting human subjectiveassessments, automatic emotional estimation is crucial. To address this issue, there is a demandfor automated emotional estimation technology. This paper provides an overview of emotionalresearch in dogs and cats, their senses, and their roles in interaction with humans. In addition, wedescribed automated emotional estimation technology using image/video and electrocardiographyas a complement to human ability for emotional recognition of dogs and cats, and its applications forimplementation. Practical implementation of automated emotional estimation technology shouldbe adaptable to different breeds, individuals, and environmental conditions. Improving this technol-ogy has the potential to contribute to various fields, including pet welfare enhancement, veterinarycare, ethology, and support for humans with dogs or cats.journal articl
Study of Emotion Concept Formation by Integrating Vision, Physiology, and Word Information using Multilayered Multimodal Latent Dirichlet Allocation
No full textHow are emotions formed? Through extensive debate and the promulgation of diverse theories , the theory of constructed emotion has become prevalent in recent research on emotions. According to this theory, an emotion concept refers to a category formed by interoceptive and exteroceptive information associated with a specific emotion. An emotion concept stores past experiences as knowledge and can predict unobserved information from acquired information. Therefore, in this study, we attempted to model the formation of emotion concepts using a constructionist approach from the perspective of the constructed emotion theory. Particularly, we constructed a model using multilayered multimodal latent Dirichlet allocation , which is a probabilistic generative model. We then trained the model for each subject using vision, physiology, and word information obtained from multiple people who experienced different visual emotion-evoking stimuli. To evaluate the model, we verified whether the formed categories matched human subjectivity and determined whether unobserved information could be predicted via categories. The verification results exceeded chance level, suggesting that emotion concept formation can be explained by the proposed model.journal articl
イヤシ タイケン ： VR / AR エージェント ニ ヨル ココチヨイ 「 ハナシ ナガラ ナデル 」 ドウサ
Get PDFロボットによる「話しながらなでる」動作は人の快感情を増加させ痛みを軽減する効果が確認されている．しかし，人間の感覚は90%以上が視覚情報であるため，どのような人が，どのように「話しながらなでる」かによって，快感情や痛み軽減効果に変化があると考えられる．本研究では，VR/AR環境で人型エージェントによる「話しながらなでる」システムを開発する．このシステムは触覚，聴覚，視覚を統合した上質な心地よい体験を提供する．conference pape
Analysis of Multi-Channel sEMG Data Using Transformer and Construction of a Simulated Abnormal Gait Dataset
Get PDFdoctoral thesi
Shape-Modified Ultrathin Glass Sheet Cantilever for Precise Single Cell Stiffness Measurement
No full textDue to its small size, label-free nature, and excellent compatibility with surrounding instruments, the microcantilever-based sensor is widely used in physical, chemical, and biological measurements. However, traditional microcantilevers face limitations due to complex fabrication methods, intricate operating procedures, or low sensitivity. In this research, we propose a shape-modified, 10 μm ultrathin glass sheet (UTGS)-based cantilever that is highly sensitive (nN/μm) and flexible, integrated with a strain gauge sensor. We present its application in measuring the mechanical properties of single cells. Compared with conventional methods, the proposed UTGS-based cantilever is easier to fabricate, offers superior physical and chemical properties, and demonstrates a high linear correlation between voltage change and applied force or displacement. The cantilever has a sensitivity of 190 nN/μm, making it suitable for measuring the cell stiffness. Additionally, it exhibits excellent optical transparency, enabling real-time observation during measurements, as demonstrated in the stiffness measurement of Dictyostelium cells.journal articl