Source : Verilog HDL

Only VLSI Implementation Cost : Rs. 40,000/-

Only MATLAB Implementation Cost : Rs. 6,000/-

Both VLSI + MATLAB Implementation Cost : Rs. 45,000/-

Proposed Abstract:

Speaker recognition is one of the technologies that may be used for biometric identification, and it offers higher application possibilities in many sectors. At the moment, the implementation of the speaker identification algorithm on the hardware is mostly dependent on the SOC of the FPGA. An FPGA-based real-time technique for extracting acoustic characteristics is presented in this research. The method is based on MFCC, which stands for Mel Frequency Cepstral Coefficients. The trials have shown that the FPGA-based MFCC calculation has a high level of accuracy; the purpose of this study is to enhance the performance assessment of MFCC by making use of novelty-based architecture. In this study, a technique for FPGA-based speech recognition is provided. This approach was developed by investigation and analysis of the speaker recognition algorithm. The IFFT, the Mell filter, the DFT, the derivatives, and the Hamming Window with pre-emphasis are every aspect of this approach. This proposed MFCC will be constructed with an AHB interface in order to facilitate higher access DMA Controller when it is used in SOC applications. This work was carried out using Verilog HDL, and it was generated with Xilinx Vivado FPGA. Additionally, all of the parameters were analyzed and compared with regard to area, latency, and power.

List of the following  materials will be included with the Downloaded Backup: