Search

Domain mismatch problem caused by speaker-unrelated feature has been a major topic in speaker recognition. In this paper, we propose an explicit disentanglement framework to unravel speaker-relevant features from speaker-unrelated features via mutual information (MI) minimization. To achieve our goal of minimizing MI between speaker-related and speaker-unrelated features, we adopt a contrastive log-ratio upper bound (CLUB), which exploits the upper bound of MI. Our framework is constructed in a 3-stage structure. First, in the front-end encoder, input speech is encoded into shared initial embedding. Next, in the decoupling block, shared initial embedding is split into separate speaker-related and speaker-unrelated embeddings. Finally, disentanglement is conducted by MI minimization in the last stage. Experiments on Far-Field Speaker Verification Challenge 2022 (FFSVC2022) demonstrate that our proposed framework is effective for disentanglement. Also, to utilize domain-unknown datasets containing numerous speakers, we pre-trained the front-end encoder with VoxCeleb datasets. We then fine-tuned the speaker embedding model in the disentanglement framework with FFSVC 2022 dataset. The experimental results show that fine-tuning with a disentanglement framework on a existing pre-trained model is valid and can further improve performance., Accepted by APSIPA ASC 2022. Camera-ready. 8 pages, 4 figures, and 1 table

th , 2015; Revised: August 11 st , 2015; Accepted: August 20 th , 2015 Purpose: From the perspective of biomechanics, joint moments quantitatively show a subject’s ability to perform actions. In this study, the effect of normalization in the fast and asymmetric motions of a golf swing was investigated by applying three different normalization methods to the raw joint moment. Methods: The study included 13 subjects with no previous history of musculoskeletal diseases. Golf swing analyses were performed with six infrared cameras and two force plates. The majority of the raw peak joint moments showed a significant correlation at p < 0.05. Additionally, the resulting effects after applying body weight (BW), body weight multiplied by height (BWH), and body weight multiplied by leg length (BWL) normalization methods were analyzed through correlation and regression analysis. Results: The BW, BWH, and BWL normalization methods normalized 8, 10, and 11 peak joint moments out of 18, respectively. The best method for normalizing the golf swing was found to be the BWL method, which showed significant statistical differences. Several raw peak joint moments showed no significant correlation with measured anthropometrics, which was considered to be related to the muscle coordination that occurs in the swing of skilled professional golfers. Conclusions: The results of this study show that the BWL normalization method can effectively remove differences due to physical characteristics in the golf swing analysis.

Permanent facial nerve damage is one of the most critical complications that can occur during otologic and neurosurgical procedures. As a result, facial nerve monitoring equipment is now used in most of otologic surgeries. The monitoring devices used for this purpose usually evaluate the activity of facial muscle in response to stimulation to the facial nerve. For these purposes, electromyography and the train-of-four (TOF) method is commonly used clinically, but these approaches have an important limitation: this qualitative method only depends on visual information of finger movement. Thus, in this study, a new factor that can be used to quantitatively the stage of muscle relaxation at the facial nerve was developed based on the degree of muscle relaxation and electro-stimulation. Ketamine intravenous anesthesia and isoflurane inhalation anesthesia was given to 13 rabbits. A 28 G unipolar needle-electrode was inserted at an exposed facial nerve, and stimulated (0.1 mA–5.0 mA) with an electrical stimulator. Also, the degree of muscle relaxation based electromyographic signal acquisition was performed while increasing the dose of neuromuscular blockades. After data noise reduction and normalization, the collected EMG signals were divided into 33 variables, and a new evaluation factor was developed using statistical methods based on multiple linear regression models and principal component analysis. A new evaluation factor was identified that can be used to quantitatively estimate each phase in muscle relation. The evaluating factors that were calculated in accordance with the electrical simulation resulted in improved results (r = 0.96) when compared to the existing assessment methods. (T1 ratio r = 0.72; T4/T1 ration r = 0.52).

Purpose: The goal of this study was to build an accurate digital factory that evaluates the performance of a factory using computer simulation. To achieve this goal, we evaluated the effect of worker-related variables on production in a simulation model using comparative analysis of two cases. Methods: The overall work process and worker-related variables were determined and used to build a simulation model. Siemens PLM Software's Plant Simulation was used to build a simulation model. Also, two simulation models were built, where the only difference was the use of the worker-related variable, and the total daily production analyzed and compared in terms of the in dividual process. Additionally, worker efficiency was evaluated based on worker analysis. Results: When the daily production of the two models were compared, a 0.16% error rate was observed for the model where the worker-related variables were applied and error rate was approximately 5.35% for the model where the worker-related variables were not applied. In addition, the production in the individual processes showed lower error rate in the model that included the worker-related variables than the model where the worker-related variables were not used. Also, among the total of 22 workers, only three workers satisfied the IFRS (International Financial Reporting Standards) suggested worker capacity rate (90%). Conclusions: In the daily total production and individual process production, the model that included the worker-related variables produced results that were closer to the real production values. This result indicates the importance of work er elements as input variables, in regards to building accurate simulation models. Also, as suggested in this study, the model that included the worker-related variables can be utilized to analyze in more detail actual production. The results from this study are expected to be utilized to improve the work process and worker efficiency.