Although state of the art in many typical machine learning tasks, deep learning algorithmsareverycostly interms ofenergyconsumption,duetotheirlargeamount of required computations and huge model sizes. Because of this, deep learning applications on battery-constrained wearables have only been possible through wireless connections with a resourceful cloud. This setup has several drawbacks. First, there are privacy concerns. Cloud computing requires users to share their raw data—images, video, locations, speech—with a remote system. Most users are not willing to do this. Second, the cloud-setup requires users to be connected all the time, which is unfeasible given current cellular coverage. Furthermore, real-time applications require low latency connections, which cannot be guaranteed using the current communication infrastructure. Finally, wireless connections are very inefficient—requiringtoo much energyper transferredbit for real-time data transfer on energy-constrained platforms.
标签： Embedded_Deep_Learning Algorithms
上传时间： 2020-06-10
上传用户：shancjb
Artificial Intelligence (AI) is a big field, and this is a big book. We have tried to explore the full breadth of the field, which encompasses logic, probability, and continuous mathematics; perception, reasoning, learning, and action; and everything from microelectronic devices to robotic planetary explorers. The book is also big because we go into some depth. The subtitle of this book is “A Modern Approach.” The intended meaning of this rather empty phrase is that we have tried to synthesize what is now known into a common frame- work, rather than trying to explain each subfield of AI in its own historical context. We apologize to those whose subfields are, as a result, less recognizable.
标签： A-Modern-Approach Intelligence
上传时间： 2020-06-10
上传用户：shancjb
This design uses Common-Emitter Amplifier (Class A) with 2N3904 Bipolar Junction Transistor. Use “Voltage Divider Biasing” to reduce the effects of varying β (= ic / ib) (by holding the Base voltage constant) Base Voltage (Vb) = Vcc * [R2 / (R1 + R2)] Use Coupling Capacitors to separate the AC signals from the DC biasing voltage (which only pass AC signals and block any DC component). Use Bypass Capacitor to maintain the Q-point stability. To determine the value of each component, first set Q-point close to the center position of the load line. (RL is the resistance of the speaker.)
上传时间： 2020-11-27
上传用户：
This edition of Digital Image Processing is a major revision of the book. As in the 1977 and 1987 editions by Gonzalez and Wintz, and the 1992, 2002, and 2008 editions by Gonzalez and Woods, this sixth-generation edition was prepared with students and instructors in mind. The principal objectives of the book continue to be to provide an introduction to basic concepts and methodologies applicable to digital image processing, and to develop a foundation that can be used as the basis for further study and research in this field. To achieve these objectives, we focused again on material that we believe is fundamental and whose scope of application is not limited to the solution of specialized problems. The mathematical complexity of the book remains at a level well within the grasp of college seniors and first-year graduate students who have introductory preparation in mathematical analysis, vectors, matrices, probability, statistics, linear systems, and computer programming. The book website provides tutorials to support readers needing a review of this background material
标签： Processing Digital Image
上传时间： 2021-02-20
上传用户：
一份极好PCB的EMC设计指南 这是一份关闭PCB方面EMC的一些讲解，里面对应PCB设计讲得比较详细，对于从事PCB设计相关的人，可以学习一下。
标签： emc
上传时间： 2022-01-28
上传用户：
这是我在做大学教授期间推荐给我学生的一本书，非常好，适合入门学习。《python深度学习》由Keras之父、现任Google人工智能研究员的弗朗索瓦•肖莱（François Chollet）执笔，详尽介绍了用Python和Keras进行深度学习的探索实践，包括计算机视觉、自然语言处理、产生式模型等应用。书中包含30多个代码示例，步骤讲解详细透彻。作者在github公布了代码，代码几乎囊括了本书所有知识点。在学习完本书后，读者将具备搭建自己的深度学习环境、建立图像识别模型、生成图像和文字等能力。但是有一个小小的遗憾：代码的解释和注释是全英文的，即使英文水平较好的朋友看起来也很吃力。本人认为，这本书和代码是初学者入门深度学习及Keras最好的工具。作者在github公布了代码，本人参照书本，对全部代码做了中文解释和注释，并下载了代码所需要的一些数据集（尤其是“猫狗大战”数据集），并对其中一些图像进行了本地化，代码全部测试通过。（请按照文件顺序运行，代码前后有部分关联）。以下代码包含了全书约80%左右的知识点，代码目录：2.1: A first look at a neural network（ 初识神经网络）3.5: Classifying movie reviews（电影评论分类：二分类问题）3.6: Classifying newswires（新闻分类：多分类问题 ）3.7: Predicting house prices（预测房价：回归问题）4.4: Underfitting and overfitting（ 过拟合与欠拟合）5.1: Introduction to convnets（卷积神经网络简介）5.2: Using convnets with small datasets（在小型数据集上从头开始训练一个卷积网络）5.3: Using a pre-trained convnet（使用预训练的卷积神经网络）5.4: Visualizing what convnets learn（卷积神经网络的可视化）
上传时间： 2022-01-30
上传用户：
Extensively rewritten to present the C++11 language, standard library, and key design techniques as an integrated whole, Stroustrup thoroughly addresses changes that make C++11 feel like a whole new language, offering definitive guidance for leveraging its improvements in performance, reliability, and clarity. C++ programmers around the world recognize Bjarne Stoustrup as the go-to expert for the absolutely authoritative and exceptionally useful information they need to write outstanding C++ programs. Now, as C++11 compilers arrive and development organizations migrate to the new standard, they know exactly where to turn once more: Stoustrup's C++ Programming Language, Fourth Edition.Bjarne Stroustrup是C++的设计师和最早的实现者，也是《C++程序设计语言》、《带标注的C++参考手册》和《C++语言的设计与演化》的作者。他从丹麦Aarhus大学和英国牛津大学毕业，现在是AT&T大规模程序设计研究部的负责人，AT&T特别成员，AT&T贝尔实验室特别成员，以及ACM特别成员。Stroustrup的研究兴趣包括分布式系统、操作系统、模拟、设计和程序设计。他也是Addison·Wesley的C++In-Depth系列书籍的编辑。
标签： C++
上传时间： 2022-02-01
上传用户：
Single chip TFT-LCD Controller/Driver with On-chip Frame Memory (FM) Display Resolution: 240*RGB (H) *320(V) Frame Memory Size: 240 x 320 x 18-bit = 1,382,400 bits LCD Driver Output Circuits- Source Outputs: 240 RGB Channels- Gate Outputs: 320 Channels- Common Electrode Output Display Colors (Color Mode)- Full Color: 262K, RGB=(666) max., Idle Mode Off- Color Reduce: 8-color, RGB=(111), Idle Mode On Programmable Pixel Color Format (Color Depth) for Various Display Data input Format- 12-bit/pixel: RGB=(444)- 16-bit/pixel: RGB=(565)- 18-bit/pixel: RGB=(666) MCU Interface- Parallel 8080-series MCU Interface (8-bit, 9-bit, 16-bit & 18-bit)- 6/16/18 RGB Interface(VSYNC, HSYNC, DOTCLK, ENABLE, DB[17:0])- Serial Peripheral Interface(SPI Interface)- VSYNC Interface
上传时间： 2022-03-04
上传用户：
本文首次设计并验证了基于macom三合一芯片设计的光模块电路,该电路旨在提供一种满足SFF-8472中规定的数字诊断功能的低成本SFP+模块。电路采用激光器驱动、限幅放大器、控制器以及时钟恢复单元集成的单芯片,在保证高精度数字诊断功能基础上,实现了低成本高可靠的特点。该电路在光接收接口组件与激光器驱动和限幅放大器单元的限幅放大器部分之间接入滤波器来提高模块的灵敏度及信号质量。在控制器单元的数字电位器的引脚上采用外加电阻的方式避免出现上电不发光的故障问题。该研究结果为下一代SFP-DD光模块设计与开发工作,奠定了一定的理论与实践基础。This paper designs and validates the optical module circuit based on the MACOM Trinity chip for the first time.This circuit aims to provide a low-cost SFP module which meets the digital diagnosis function specified in SFF-8472.The circuit uses a single chip integrated with laser driver,limiting amplifier,controller and clock recovery unit.On the basis of ensuring high precision digital diagnosis function,it achieves the characteristics of low cost and high reliability.The circuit connects a filter between the optical receiving interface module and the limiting amplifier part of the laser driver and limiting amplifier unit to improve the sensitivity and signal quality of the module.The pin of the digital potentiometer in the controller unit is equipped with an external resistance to avoid the problem of power failure.The research results lay a theoretical and practical foundation for optical module design in high-speed data center.
上传时间： 2022-04-03
上传用户：
《HeadFirstJava》是一本完整地面向对象(object-oriented，OO)程序设计和Java的学习指导用书，根据学习理论所设计，你可以从程序语言的基础开始，到线程、网络与分布式程序等项目。重要的是，你可以学会如何像一个面向对象开发者一样去思考，而且不只是读死书。 在这里，你可以会玩游戏、拼图、解谜题以及以意想不到的方式与Java交互。 在这些活动中，你还会写出一堆真正的Java程序，如一个船舰炮战游戏和一个网络聊天程序等等。 “HeadFirst系列”图文并茂学习方式能让你快速地在脑海中掌握住知识，敞开心胸准备好学习这些关键性的主题： ★Java程序语言 ★面向对象程序开发 ★Swing图形化接口 ★使用JavaAPI函数库 ★编写、测试与布署应用程序 ★处理异常；多线程 ★网络程序设计 ★集合与泛型
标签： java
上传时间： 2022-06-12
上传用户：
