Estimating multiple pitch frequencies of concurrent speech sources from a single-microphone input is essential to speech separation. Nevertheless, pitch cues of individual sources are weakened by each other, making the estimation unreliable. This paper presents a pitch tracking method that incorporated in a model-based separation framework. Multiple pitch estimation is simplified into single pitch estimation by segregating the source envelope from mixture spectrum with statistics of familiar speech patterns. Comprehensive experiments have compared the proposed tracking method with a recently reported multiple pitch estimator and its modified version equipped with ideal pitch cues. Lower estimation errors are achieved. Furthermore, this approach is applicable to other model-based frameworks as well.
A ortocaval compression (ACC) in parturients occurs when the uterus compresses the abdominal aorta and inferior vena cava (IVC). This compression can lead to maternal hypotension and uteroplacental hypoperfusion. However, potentially significant ACC is difficult to detect before a regional block because many patients have no signs or symptoms. Thus, predicting the patient in whom profound hypotension may occur after sympathectomy during regional anesthesia is difficult as well. A noninvasive technique for identifying patients with significant ACC would be helpful as invasive techniques, such as radiologic angiography, are not practical for routine clinical use. This prospective observational study was designed to measure changes in cardiac output (CO) and other maternal hemodynamic values at different degrees of lateral tilt. The authors hypothesized that this would allow them to identify the presence of IVC compression in term parturients.The authors recruited nonlaboring parturients at term who were to undergo elective cesarean section. Standard monitoring was applied, including noninvasive arterial pressure monitoring at 1-minute intervals on the left arm, ECG, pulse oximetry, and continuous cardiotocography. A second noninvasive arterial pressure monitor was placed on the patient's left calf for measuring lower limb arterial pressure (AP). Intermittent measurements of CO, stroke volume, and systemic vascular resistance were obtained using suprasternal Doppler ultrasound. Patient hemodynamic measurements were made before spinal anesthesia with the operating table placed at 4 levels of left lateral tilt sequentially applied: 0 degrees (lying completely supine), 7.5 degrees, 15 degrees, and 90 degrees (complete left lateral with the hips and knees slightly flexed). Patients were kept in each tilted position for at least 5 minutes to allow for stabilization of the parameters before hemodynamic measurements were taken. A Z20 mm Hg difference in CO was thought to be evidence of significant IVC compression and a Z20 mm Hg difference in AP between the upper and lower limb was thought to reflect aortic compression.Of 170 patients who provided consent, 157 patients had complete data sufficient for analysis. No patients had episodes of hypotension or fetal heart rate abnormalities during the study period. CO was on an average 5% higher when patients were positioned at 15-and 90-degree tilt compared with 0-and 7.5-degree tilt. CO values at the 0, 7.5, 15, and 90-degree positions were 5.9, 5.9, 6.2, and 6.3 L/min, respectively (P = 0.001 for comparisons between 15-and 90-degree tilt compared to 0-and 7.5-degree tilt). Stroke volume values were 74, 74, 76, and 78 mL, respectively (no difference detected) and systemic vascular resistances were 1006, 1024, 934, and 979, respectively (P = 0.003, for comparisons between 15-and 90-degree tilt compared with 0-and 7.5-degree tilt). Heart rates were 81, 80, 80, and 82 bpm, respectively (no difference). In 11 patients who had a Z20% difference in CO between tilted positions, the me...
In standard speech recognition systems in which training data are clean speech, the presence of background noise in received signal can severely deteriorate the recognition performance. This paper presents a simple noise-robust speech recognition system based on a modified noise spectral estimation method called Mainlobe-Resilient Time-Frequency Quantile-based Noise Estimation (M-R T-F QBNE), which focuses on the mainlobes at harmonic frequencies. We estimate the global signal-to-noise ratio (SNR) and select a recognition model, which is best matched to the SNR operating range. Experimental results show that the recognition accuracy of the proposed recognition system is higher than that of the AURORA2 clean training baseline by 23%. Compared to multicondition training, the proposed method achieves comparable recognition accuracy. 1
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
customersupport@researchsolutions.com
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.