Bug fixes and perf improvements in 0.4

• Improved performance of guided backprop with subsequent calls.
• various minor bug/doc fixes.

guided backprop, API improvements.

• Introducing grad_modifiers, backprop_modifiers, input_modifiers.
• Implemented guided and rectified backprop.
• API changes. Deprecated visualize_regression variant and removed RegressionTarget loss,
• Added visualize_xxx_with_losses variant for flexibility.
• Moved all examples to notebooks.
• Introduced backend abstraction for backend specific code.
• Modularized visualization code.

Bug fixes

Fixes various minor issues in 0.3.2

Introducing regression visualizations.

• Implemented regression variant for all visualizations.
• Refactored common code for various visualizations for advanced use cases of specifying custom loss.
• Added RegressionTarget experimental loss function.
• Added an example use-case for visualizing a toy self-driving car.
• Added the ability to specify input range for activation visualizations
• Decoupled heatmap overlay in attention visualizations as not all inputs are images.

Various bug fixes

0.3.1

Update PyPI

N-dim support

• Added support for N-dim attention, and activation maximization
• Huge perf improvements
• Removed a ton of unwanted dependencies.

Added attention maps

API changes, added attention related visualizations and examples.

Initial Release

Supports

• Conv/Dense layer guided backprop
• Saliency maps
• grad-CAM