Skip to main content

> ML_LITERATURE // CHOLLET-2017-XCEPTION-DEEP-LEARNING-DEPTHWISE-SEPARABLE-CONVOLUTIONS_v1.0

Xception: Deep Learning with Depthwise Separable Convolutions

François Chollet · IEEE Conference on Computer Vision and Pattern Recognition (CVPR) (2017)

seminal-architecture2017foundationalthirdPartyReproduced
DOI: 10.1109/CVPR.2017.195Back to Literature

Principal Contribution

Hypothesized that cross-channel correlations and spatial correlations can be entirely decoupled, introducing depthwise separable convolutions.

Operational Relevance

Serves as qualified reference for deploying task-image-classification in production.

Assumptions

  • Spatial feature coherence and data manifold structure adhere to continuous representation hypotheses

Limitations

  • Computational complexity scales with spatial resolution and parameter capacity

Connected Algorithms, Architectures & Tools

Related Algorithms:
Related Architectures:
Implementing Libraries: