AI for Image Processing
Understand and implement state-of-the-art image-processing models (CNN, GAN, Auto-encoder) with TensorFlow 2 / Keras.
Get Course Info
Audience: Developers · Data Analysts · Data Scientists
Duration: 3 days
Format: Lectures & hands-on labs (50 % / 50 %)
Overview
AI algorithms for image analysis have made tremendous progress thanks to abundant data, affordable compute, and libraries such as TensorFlow. This course introduces Deep-Learning concepts plus TensorFlow and Keras, then dives into CNNs, GANs, Auto-encoders, and Transfer-Learning for computer-vision use-cases.
Objective
Understand and implement state-of-the-art image-processing models (CNN, GAN, Auto-encoder) with TensorFlow 2 / Keras.
What You Will Learn
- Deep-Learning concepts
- TensorFlow 2 and Keras APIs
- Building neural networks & using TensorBoard
- Deep Neural Networks
- Convolutional Neural Networks (CNN)
- Generative Adversarial Networks (GAN)
- Auto-encoders
- Transfer-Learning & benchmarking on CPU/GPU
Course Details
Audience: Developers · Data Analysts · Data Scientists
Duration: 3 days
Format: Lectures & hands-on labs (50 % / 50 %)
- Basic Python & Jupyter notebooks
Setup: Cloud-based lab (zero install) · Modern laptop · Chrome
Detailed Outline
- AI / ML / DL landscape
- Hardware & software ecosystem
- Supervised / Unsupervised / Reinforcement learning
- Execution graph
- GPU / TPU support
- TensorFlow API
- Lab: setup & run TensorFlow
- Models & Layers
- Using Keras API
- Lab
- Perceptrons
- Activation functions
- Back-propagation
- Optimisers (GD, Adam, RMSProp)
- Loss functions
- Vanishing / Exploding gradients
- Lab: TF playground
- Architecture & sizing
- Lab: FFNN
- CNN architecture & concepts
- Lab: image recognition
- GAN overview
- Generating images
- Lab: GAN
- Use-cases & architecture
- Lab: auto-encoder
- Customising pre-trained models
- Lab: transfer-learning & benchmarking
- Group project on a real computer-vision use-case
Ready to Get Started?
Contact us to learn more about this course and schedule your training.