The 4.0 kbit/s speech codec described in this paper is based on a Frequency Domain Interpolative (FDI) coding technique, which belongs to the class of prototype waveform Interpolation (PWI) coding techniques. The codec also has an integrated voice activity detector (VAD) and a noise reduction capability. The input signal is subjected to LPC analysis and the prediction residual is separated into a slowly evolving waveform (SEW) and a rapidly evolving waveform (REW) components. The SEW magnitude component is quantized using a hierarchical predictive vector quantization approach. The REW magnitude is quantized using a gain and a sub-band based shape. SEW and REW phases are derived at the decoder using a phase model, based on a transmitted measure of voice periodicity. The spectral (LSP) parameters are quantized using a combination of scalar and vector quantizers. The 4.0 kbits/s coder has an algorithmic delay of 60 ms and an estimated floating point complexity of 21.5 MIPS. The performance of this coder has been evaluated using in-house MOS tests under various conditions such as background noise. channel errors, self-tandem. and DTX mode of operation, and has been shown to be statistically equivalent to ITU-T (3.729 8 kbps codec across all conditions tested.
標簽： frequency-domain interpolation performance Design kbit_s speech coder based and of
上傳時間： 2018-04-08
上傳用戶：kilohorse
This book provides the essential design techniques for radio systems that operate at frequencies of 3 MHz to 100 GHz and which will be employed in the telecommunication service. We may also call these wireless systems, wireless being synonymous with radio, Telecommunications is a vibrant indus- try, particularly on the ‘‘radio side of the house.’’ The major supporter of this upsurge in radio has been the IEEE and its 802 committees. We now devote ? . an entire chapter to wireless LANs WLANs detailed in IEEE 802.11. We also now have subsections on IEEE 802.15, 802.16, 802.20 and the wireless ? . ? metropolitan area network WMAN . WiFi, WiMax,, and UWB ultra wide- . band are described where these comparatively new radio specialties are demonstrating spectacular growth.
標簽： Telecommunication Design System Radio for
上傳時間： 2020-06-01
上傳用戶：shancjb
In 2001, Orange, a UK mobile network operator, announced the “Orange at Home” project, a smart house incorporating the latest technology wizardry built some 20 miles north of London. It was intended to be more than a mere showcase, with plans for real families to move in and live with the smart home. My then research establishment, the Digital World Research Centre at the University of Surrey, was commissioned to study how these families reacted to their new home, and to report lessons for the future development of smart homes and smart home technologies.
上傳時間： 2020-06-06
上傳用戶：shancjb
Over the past few decades there has been an exponential growth in service robots and smart home technologies, which has led to the development of exciting new products in our daily lives. Service robots can be used to provide domestic aid for the elderly and disabled, serving various functions ranging from cleaning to enter- tainment. Service robots are divided by functions, such as personal robots, field robots, security robots, healthcare robots, medical robots, rehabilitation robots and entertainment robots. A smart home appears “intelligent” because its embedded computers can monitor so many aspects of the daily lives of householders. For example, the refrigerator may be able to monitor its contents, suggest healthy alter- natives and order groceries. Also, the smart home system may be able to clean the house and water the plants.
標簽： Robotics Service Digital within Home the
上傳時間： 2020-06-06
上傳用戶：shancjb
這是我在做大學教授期間推薦給我學生的一本書，非常好，適合入門學習。《python深度學習》由Keras之父、現任Google人工智能研究員的弗朗索瓦?肖萊（Franc?ois Chollet）執筆，詳盡介紹了用Python和Keras進行深度學習的探索實踐，包括計算機視覺、自然語言處理、產生式模型等應用。書中包含30多個代碼示例，步驟講解詳細透徹。作者在github公布了代碼，代碼幾乎囊括了本書所有知識點。在學習完本書后，讀者將具備搭建自己的深度學習環境、建立圖像識別模型、生成圖像和文字等能力。但是有一個小小的遺憾：代碼的解釋和注釋是全英文的，即使英文水平較好的朋友看起來也很吃力。本人認為，這本書和代碼是初學者入門深度學習及Keras最好的工具。作者在github公布了代碼，本人參照書本，對全部代碼做了中文解釋和注釋，并下載了代碼所需要的一些數據集（尤其是“貓狗大戰”數據集），并對其中一些圖像進行了本地化，代碼全部測試通過。（請按照文件順序運行，代碼前后有部分關聯）。以下代碼包含了全書約80%左右的知識點，代碼目錄：2.1: A first look at a neural network（ 初識神經網絡）3.5: Classifying movie reviews（電影評論分類：二分類問題）3.6: Classifying newswires（新聞分類：多分類問題 ）3.7: Predicting house prices（預測房價：回歸問題）4.4: Underfitting and overfitting（ 過擬合與欠擬合）5.1: Introduction to convnets（卷積神經網絡簡介）5.2: Using convnets with small datasets（在小型數據集上從頭開始訓練一個卷積網絡）5.3: Using a pre-trained convnet（使用預訓練的卷積神經網絡）5.4: Visualizing what convnets learn（卷積神經網絡的可視化）
上傳時間： 2022-01-30
上傳用戶：
