Instructors: Dr. Kiran Gunnam
What you will learn / Topics that will be covered:
- Intuitive Treatment
- Overview of Deep Neural Networks and Deep Generative Models – History and early Neural Networks, A visual proof that neural nets can compute any function, Convnet (LeNet), AlexNet, Recurrent Neural Networks, Deep Belief Networks, Autoencoder and Generative Adversarial Networks
- Convolutional Layer in Details– Building Block of CNN, Convolutional layer, CNN Architecture Activation functions, Pooling, Normalization, Fully connected layers, Soft-max, Training/Back propagation.
- CNN Architecture Optimization in Detail– Overfitting,Data Augmentation, Dropouts, Early stopping, Transfer Learning, Adversarial Training, Effect of Depth, Layer Patterns, Layer Sizing, Kernel size, stride length
- CNN Visualization
- Case studies of CNN Architectures – LeNet, AlexNet, ZFNet, VGGNet, GoogleNet, Residual Net, MobileNets
- Summary and Cautionary words
- In-depth Treatment
- VGGNet and Network Pruning
- Inception V4
- Progressive Neural Architecture Search (PNASNet)
- Generative Adversarial Networks (GANs)
- Hands-on Practice
- CNN – MNIST (handwritten digits recognition) with Tensorlfow
- MobileNet – Transfer Learning for Image Classification with Tensorflow
Engineers, researchers, practitioners and students who are interested in machine learning, deep learning, convolutional neural networks, and their implementations on GPUs. This workshop will particularly benefit people who intend to develop deep learning techniques using CNN and applications that can keep improving themselves after seeing more and diverse data to achieve intelligence.
Basic knowledge of Machine Learning, and familiarity with Python and Tensorflow basics
Upon completion of this course, you’ll be able to start solving problems using Deep Learning with CNN such as image classification and will be ready to apply the techniques to real-world problems such as perception for autonomous vehicles.