What is Deep Learning?
Deep learning is a new class of artificial intelligence. Deep learning is the past, present, and future of AI as it has many impressive successes and it is growing market worldwide. It uses the processing power of neural networks to create images and to interpret and learn from them. The technique in DL is used to train more powerful image recognition systems, and also to enhance the performance of other algorithms.
How Deep Learning works?
Deep learning methods combine multiple layers of neural networks to learn complex models of images in order to interpret them. Each neural network in a DL is connected to other neural networks and data. It provides the computational resources needed to process the input.
Why we say that “THE PAST, PRESENT, AND FUTURE OF AI: DEEP LEARNING”?
Learning involves:
- processing of input data for e.g. a video frame or an image
- training a particular model to perform a task on the data
- evaluating the model to improve the performance of the task
In DL, layers are made of neurons connected to other layers. Each layer having an output which is processed by a layer below it. In this fashion, a single image learns and processes multiple times. It allows the model to train & to perform many different tasks. So, we say that THE PAST, PRESENT, AND FUTURE OF AI: DEEP LEARNING.
The simplest type of neural network is known as a multi-layer perceptron. Each layer consists of one neuron.
The first layer is to train the system how to classify images into categories, such as cat or dog. By doing this several times, we can then learn how to do all the other tasks. For e.g. recognizing human faces, facial expressions and other types of information in photos.
Applications of Deep Learning in Present world-
As you know, facebook is capable of recognizing human faces and giving information on the basis of a photo detected. It can recognize the image of a person even if the photo is blurr.
Deep learning goes beyond facial recognition. It can also be used for:
- Identifying people from their tattoos
- Recognizing people by their facial features
- Knowing name of a person from his/her voice
- Associating people by their handwriting
Some of the machine learning applications are:
- Automatic speech recognition
- Image recognition
- Video and image processing
- Natural language processing
