94 GHz Millimeter Wave Conducted Speech Enhancement
Abstract: The traditional
microphone speech acquisition method has several limitations, such as short acquisition
distance, non-directivity, weak ability to compensate for anti-acoustic noise,
and poor sensitivity. As an alternative, our lab developed a new type of speech
detection system using bio-radar. This novel speech detection method may
effectively compensate for the shortcomings of the traditional microphone
because of the special features of bio-radar. However, the detected radar
speech is also corrupted by specific electromagnetic and ambient noises.
Therefore, this study proposes a minimum mean-square error (MMSE) short-time
spectral amplitude estimate algorithm to enhance radar captured speech. Because
the proposed approach employs an optimal spectral amplitude estimator as well
as a method to minimize the mean-square error of the log-spectra in enhancing
noisy speech, this algorithm significantly reduces the colored residual noise.
The enhanced radar speech was evaluated by both subjective and objective
quality evaluation methods. The results demonstrate that the bio-radar can successfully
detect speech signal and the proposed speech enhancement algorithm can improve
radar captured speech quality.
Author: S. Li
Journal: jptkomputergg170176
Artikel Terkait :
Jp Teknik Komputer gg 2017
- Social Media Success Model for Knowledge Sharing (Scale Development and Validation)
- Histogram Equalization for Improving Quality of Low-Resolution Ultrasonography Images
- Research and Application of Development Model of Information Service for IOT of Oil and Gas Production Based on Cloud Architecture
- A Customized Reconfiguration Controller with Remote Direct ICAP Access for Dynamically Reconfigurable Platform
- Design LTE Microstrip Antenna Rectangular Patch with Beetle-Shaped Slot
- Design of Circular Patch with Double C-Shaped Slot Microstrip Antenna for LTE 1800 MHz
- A Three-Phase Grid-Connected PV System Based on SAPF for Power Quality Improvement
- Direct Torque Control: Stator Flux Regulation Improvement at Low Speed
- Brown’s Weighted Exponential Moving Average Implementation in Forex Forecasting
- A Crop Pests Image Classification Algorithm Based on Deep Convolutional Neural Network
- Hierarchical Gaussian Scale-Space on Androgenic Hair Pattern Recognition
- The Pessimistic Investor Sentiments Indicator in Social Networks
- Automatic Image Annotation Using CMRM with Scene Information
- Managers Perceptions towards the Success of E-performance Reporting System
- Binarization of Ancient Document Images based on Multipeak Histogram Assumption
- Automatic Summarization in Chinese Product Reviews
- Progressive Mining of Sequential Patterns Based on Single Constraint
- A Novel Space-time Discontinuous Galerkin Method for Solving of One-dimensional Electromagnetic Wave Propagations
- Development of Ammonia Gas Leak Detection and Location Method
- Compact Power Divider Integrated with Coupler and Microstrip Cavity Filter for X-band Surveillance Radar System
- Performance of Groundplane Shaping in Four-Element Dualband MIMO Antenna
- A Mamdani Fuzzy Model to Choose Eligible Student Entry
- Potential Usage Estimation of Ground Water using Spatial Association Rule Mining
- Investigation Study of Three-Level Cascaded H-bridge Multilevel Inverter