frontiers-in-plant-science-2020a

Materials for : https://www.frontiersin.org/articles/10.3389/fpls.2020.01015/full

The two software applications mentioned in the above paper are provided here:

• XyloRef: for data collection, and
• XyloInf: for model deployment.

We provide one trained model each for wood and charcoal identification. In the paper (and supplementary information) we report five fold cross validation results for five different CNN weight initializations i.e. 25 models each for wood and charcoal. The trained weights for all these models is not provided in the spirit of good GitHub citizenship. These models can be obtained by emailing the corresponding author.

A XyloTron system is required to run these applications. Limited support maybe available from the authors for installation/operation of the software. This is research software and is provided as-is. Use the software at your own risk.