Researcher Francesco Versaci of the CRS4 Visual and Data-Intensive Computing group won the Best Paper Award with the contribution "WAVETF: A FAST 2D WAVELET TRANSFORM FOR MACHINE LEARNING IN KERAS" at the international workshop COMPUTATIONAL ASPECTS OF DEEP LEARNING - CADL 2020 (https://ailb-web.ing.unimore.it/cadl2020/) co-located with the 25th International Conference on Pattern Recognition - ICPR 2020.
The article discusses the features and performance of the WaveTF wavelet library that leverages TensorFlow to exploit GPU parallelism and can be used to enhance machine learning in existing workflows. The work originated during the initial phase of design and optimisation of neural networks for tumour tissue recognition as part of the DEEPHEALTH project (https://deephealth-project.eu/) funded by the European Union Horizon 2020 framework programme (grant No. 825111).