Recognizing and overcoming context dependency in the application of a machine learning tool for counting stomata in Setaria versus maize