[Netease Smart News April 30 news] Remember childhood, how did you start learning to identify fruits, animals, cars and other things?
Over the years, our brains have been trained to recognize these images and then further divide them into apples, oranges, bananas, cats, dogs, and horses. In addition to learning to identify food and animals, we also learned about brands and their differences: Toyota, Honda, BMW, and others.
Inspired by the biological learning process of the human brain, scientists have developed Artificial Neural Networks (ANN). "Deep learning" refers to an artificial neural network consisting of many layers of network layers. It is the fastest growing field in machine learning. It uses a deep neural network (DNN) including multiple layers of network layers to learn characterization and abstraction to understand data such as images, sounds, and text.
So how deep is the depth neural network?
Why is deep learning called "deep" learning? This is because of the structure of these ANN networks. Decades ago, the depth of the neural network was only two layers because of insufficient computing power to build a larger network. Now, more than 10 layers or more than 100 layers of neural networks exist.
Using a multi-layered network layer in deep learning, machines now have the ability to observe, learn and respond to complex situations, sometimes even better than humans do.
Typically, data scientists spend a lot of time on data preparation, feature extraction or variable selection (choose variables useful for predictive analysis). Deep learning can automate this task and make life easier.
In order to promote the development of deep learning, many technology companies have opened up their deep learning resources, such as Google's Tensorflow and Facebook's open source module Torch. Amazon released DSSTNE on GitHub, and Microsoft also released its open source deep learning toolkit CNTK on GitHub.
So today we have seen many examples of deep learning, including:
Google Translate uses deep learning and image recognition to translate speech and written language
CamFind uses mobile visual search technology to tell you what's in the picture. You only need to take a picture of the object without typing. CamFind provides fast, accurate results.
Currently, all intelligent voice assistants such as Siri, Cortana, Alexa, and Google are using deep learning for natural language processing and speech recognition.
Amazon, Netflix, and Spotify also use deep learning in their recommendation engine, and the machine recommends for you the next best movie, movie, or music.
Google PlaNet can view photos and tell the user where the photo was taken.
DCGAN is used to enhance and supplement face images.
DeepStereo: Converts a still image shot from Street View to a 3D space. By calculating the depth and color of each pixel, different perspectives can be displayed from different perspectives.
DeepMind's WaveNet can imitate human voice to generate speech, and the generated speech is more natural than the existing text-to-speech conversion system.
Paypal is using deep learning to prevent payment fraud.
So far, deep learning has helped image classification, language translation, speech recognition, and can be used to solve pattern recognition problems. There is no doubt that this is a disruptive digital technology, more and more companies are using this technology to create new business models.
(English source /readwrite compiler / machine reviser / Claire)
燑br>
Water-cooled capacitor is supercapacitor is a capacitor with a capacity of thousands of farads.According to the principle of capacitor, capacitance depends on the distance between the electrode and electrode surface area, in order to get such a large capacitance, as far as possible to narrow the distance between the super capacitor electrode, electrode surface area increased, therefore, through the theory of electric double layer and porous activated carbon electrode.
Water-Cooled Capacitor,Water-Cooled Power Capacitor,Water-Cooled Electric Heat Capacitor,Water-Cooled Electric Heating Capacitor
YANGZHOU POSITIONING TECH CO., LTD , https://www.yzpstcc.com