Extraordinary Savings! 🥳 25% off all PDF Certs & Diplomas Ends in : : :

Claim Your Discount!

Diploma in Convolutional Neural Networks in Computer Vision

In this free course, learn about the application of convolutional neural networks to various computer vision tasks.

Publisher: NPTEL
Convolutional neural networks form the crux of most sophisticated computer vision applications such as auto-tagging on Facebook and facial security features. Are you curious to learn about the software behind popular technologies such as Siri and Google Translate? This free online course will satisfy your curiosity by explaining the features of neural networks in processing sequences such as text, sound, videos and images.
Diploma in Convolutional Neural Networks in Computer Vision
  • Duration

    10-15 Hours
  • Students

  • Accreditation


Share This Course And
Earn Money  

Become an Affiliate Member





View course modules


Convolutional neural networks (CNNs) is one of the most significant breakthroughs in computer vision. The course introduces the fundamental operations and parameters of convolution. You will discover the significance of CNNs in overcoming the feedforward network’s challenges in filtering visual imagery. Following this, the application of the backpropagation algorithm across the various layers of CNNs is explained. In addition to this, the various functions of CNN architectures for extending the support beyond residual neural networks are described. The multiple approaches for understanding and visualizing CNNs are discussed. You will investigate the procedure for visualizing layer activations, retrieving images that maximally activate a neuron, and occluding parts of the image. This course explains the methods for interpreting and understanding CNN applications in analyzing images, and the prospects of the various features for discriminative localization.

Next, the course explains how to improve visual explanations for deep convolutional networks using gradient-weighted class activation maps. You will discover how deep lift and integrated gradients methods overcome saturating gradients and the necessity for using the XRAI method for better interpretability in terms of visual coherence. Following this, you will be taught how object detection was performed in the pre-deep learning era and how basic CNNs can be adapted to object detection. In addition to this, the segmentation of images using CNN architectures is explained. This process will include the linking of each pixel in an image to a class label using semantic image segmentation techniques. The concept of understanding and processing tasks for face recognition and verification are disclosed. You will explore some of the recent efforts that have used CNNs to perform face recognition tasks such as identification and verification.

Finally, you will determine how recurrent neural networks (RNNs) can perform the same task from the output of previous data, which will include the process of training RNNs using the backpropagation algorithm. The process of solving the vanishing gradient problem using changes in the architecture of an RNN is explained. You will discover how the vanishing gradient problem gave rise to the development of popular models such as gated recurrent units and long short-term memory. Finally, you will be taught about the method of processing videos using convolutional and recurrent neural networks. This is an enlightening course that will interest those studying computer science or those interested in these topics. Why wait? Sign up today and start learning about a class of artificial neural networks that have become dominant in various computer vision tasks.

Start Course Now