Computer Vision
Build CNNs with TensorFlow & Keras - free
A free Kaggle Learn course that teaches you to build convolutional neural networks with TensorFlow and Keras. Master the core components of convnets: convolution, ReLU, maximum pooling, sliding windows, custom architectures, and data augmentation. Includes bonus lessons on TPU submissions and transfer learning.
Data ScienceFreeintermediate4 hours
Submitted by @kiraRoadmap (6 steps)
1
The Convolutional Classifier
40 min#CNN Architecture#Keras Classifier#Feature Extraction#Dense Layers
2
Convolution and ReLU
40 min#Convolution Operation#Filters & Kernels#ReLU Activation#Feature Maps
3
Maximum Pooling
40 min#Pooling Layers#Downsampling#Stride & Padding#Spatial Hierarchy
4
The Sliding Window
40 min#Window Size#Stride Parameter#Padding Modes#Output Shapes
5
Custom Convnets
40 min#Architecture Design#Layer Stacks#Hyperparameters#Model Compilation
6
Data Augmentation
40 min#Augmentation Techniques#Keras Preprocessing#Training with Augmentation#Overfitting Prevention