An RNN-based robust signal bias removal (RRSBR) method is proposed for improving both the recognition performance and the computational efficiency of the SBR method fbr adverse Mandarin speech recognition. It differs ...
详细信息
An RNN-based robust signal bias removal (RRSBR) method is proposed for improving both the recognition performance and the computational efficiency of the SBR method fbr adverse Mandarin speech recognition. It differs from the SBR method in using three broad-class sub-codebooks to encode the feature vector of each frame and combining the three encoding residuals to form the frame-level signal bias estimate. A novel approach involving softly combining the board-class encoding residuals using dynamic weighting functions generated by an RNN is applied. Experimental results show that the RRSBR method significantly outperforms the SBR method.
The diffractive optical elements (DOEs) based on basic diffraction theorem have the advantages of small size, light weight, flexibility and ease of duplication. In this thesis, we use Iterative Fourier Transform algor...
详细信息
The diffractive optical elements (DOEs) based on basic diffraction theorem have the advantages of small size, light weight, flexibility and ease of duplication. In this thesis, we use Iterative Fourier Transform algorithm (IFT) to design a sample element. After that we make use of the Single Point Modify algorithm (SPM) and dynamic weighting functions to design a greyscale DOE. We also clearly explained in detail the fabrication procedures to make our DOE samples. Latter on, we use different equipments and optical systems to measure the geometry profiles and optical performance of our DOEs.
暂无评论