Boxsers

- All data augmentation methods optimized
- Addition of a new data augmentation method
- Rubberband baseline correction method added to preprocessing.py module.
- A additionnal jupyter notebook detailing the application of the SpectroCNN model and its main features
- Additional features & improvements.

- Grad-CAM method was added to the SpectroCNN model for the identification of relevant spectral elements

- Minor issues fixed & improvements mainly related to the Spectro CNN model

Coming soon in the 1.2.0 release:
- A jupyter notebook detailing the application of the SpectroCNN model and its main features
- Additional features & improvements

Minor issues fixed

New validation_metrics.py module added

Visual darktheme ++ added

intensity_normalization to spectral_normalization with some extra features like Standard Normal Variate method

**Full Changelog**: https://github.com/ALebrun-108/BoxSERS/compare/v1.0.2...v1.0.3

New
- Bug fix
- Fig. cleanup
- New logo

**Full Changelog**: https://github.com/ALebrun-108/BoxSERS/compare/v1.0.1...v1.0.2