查詢結果分析
來源資料
相關文獻
- 類神經網路語者確認系統
- 基於鑑別式自編碼解碼器之錄音回放攻擊偵測系統
- Neural Network Procedures for Taguchi's Dynamic Problems
- A Fast and Efficient Competitive Learning Design Algorithm Based on Weight Vector Training in Transform Domain
- 專家系統振動訊號圖型判別之研究
- 反傳遞模糊類神經網路於流量推估之應用
- 類神經網路(Neural Networks)的種類及其在影像處理上的應用
- C++Fuzzy類神經網路物件導向發展系統之建立
- 臺灣汽保費率之估計--對數線性費率模式與類神經網路之比較
- 運用類神經網路於股價指數之套利--以日經225指數為例
頁籤選單縮合
題 名 | 類神經網路語者確認系統 |
---|---|
作 者 | 周義昌; | 書刊名 | 電信研究 |
卷 期 | 22:3 1992.06[民81.06] |
頁 次 | 頁363-370 |
專 輯 | 類神經網路專集 |
分類號 | 312.2 |
關鍵詞 | 語者; 確認系統; 類神經網路; |
語 文 | 中文(Chinese) |
中文摘要 | 本篇論文提丑一個以類神經網路為基礎的語者確認系統。類神經網路異於傳統的確認方式,它將我們所給它的資料拿來將空間分割成數個區域。本文即是揉合了類神經網路的劃分群集能力及語音特徵的統計結果來完成一個語者確認系統。每個輸入到此系統的語音信號都被細分為數個小音框(Frame),然後對每個音框取其線性預估係數(Linear Predictive Coefficient: LPC)來當做此音框的特徵。接下來我們用雙層認知網路(Two-Layer Perceptron)來做為訓練及確認的工具。實驗結果顯示,訓練後的系統,其內部族群測試的確認正確率是99.1%,外部族群測試的確認正確率是98.3%。 |
英文摘要 | This paper proposes a speaker verification system based on neural network. Different with the reference template matching way of traditional method, the neural network can support a capability of automatically constructing the separation space from the given data. In this paper, we apply the clustering power of neural network and the statistics of the speech’s features to the speaker verification system. A speech impute is first segmented into small frames, and then we take linear predictive coefficients (LPC) as their features. Finally, we use a two-loyer perception as training and recalling machine. We collect the speech data from eight persons. Each of them spoke ten sentences. After training, the results of recalling test show that it can get very good performance (inside group test 99.1% and outside group test 98.3%). |
本系統中英文摘要資訊取自各篇刊載內容。