Glossary term
Glossary term
Architecture
A convolutional neural network architecture based on Inception, but where Inception modules are replaced with depthwise separable convolutions. Also known as Xception.
A depthwise separable convolution (also abbreviated as separable convolution) factors a standard 3D convolution into two separate convolution operations that are more computationally efficient: first, a depthwise convolution, with a depth of 1 (n ✕ n ✕ 1), and then second, a pointwise convolution, with length and width of 1 (1 ✕ 1 ✕ n).
To learn more, see Xception: Deep Learning with Depthwise Separable Convolutions.
Created for this library
A mobile app team uses a depthwise separable CNN like MobileNet for its on-device classifier so the model meets the device's compute budget.
An IoT vendor uses a depthwise separable CNN on its low-power chip to run continuous object detection without draining the battery.
A consumer electronics maker uses a depthwise separable CNN in its smart camera so person detection runs locally with minimal latency.
Definition source: Google for Developers Machine Learning Glossary | Creative Commons Attribution 4.0 License