From the course: Computer Vision for Data Scientists
Join today to access over 24,900 courses taught by industry experts.
Convolutional layers
From the course: Computer Vision for Data Scientists
Convolutional layers
- [Instructor] Convolutional layers form the backbone of convolutional neural networks. They're specifically designed to capture spatial features in input data like images. Let's break down the essential concepts and walk away with the detailed understanding of convolutional layers. Filters are a collection of small matrices with trainable weights that the network learns during training. Each filter is responsible for detecting a specific feature in the input data. The depth of the feature should match the depth of the input data, meaning a three-channel, red, green, blue image, requires a filter with three channels. Each channel in a filter is referred to as a kernel. As the network trains, the filter learns to detect more complex and higher-level features, forming a hierarchy of learned features. Combining multiple filters allows the network to capture various features simultaneously. Note that three-by-three…
Practice while you learn with exercise files
Download the files the instructor uses to teach the course. Follow along and learn by watching, listening and practicing.
Download courses and learn on the go
Watch courses on your mobile device without an internet connection. Download courses using your iOS or Android LinkedIn Learning app.