Table 1 Summary of plant functional trait datasets. Information on the maximum number of images per species, total number of species (\(\mathrm {N_{species}}\)), number of images with woody (\(\mathrm {N_{woody}}\)) and non-woody species (\(\mathrm {N_{non-woody}}\)), number of observations with non-zero standard deviations for Plasticity setup (\(\mathrm {N_{TA}}\)), number of images in training (\(\mathrm {N_{training}}\)), validation (\(\mathrm {N_{validation}}\)) and test (\(\mathrm {N_{test}}\)) dataset as well as total number of images (\(\mathrm {N_{total}}\)) in each trait dataset. LA, leaf area; GH, growth height; SLA, specific leaf area; LNC, leaf nitrogen concentration; SM, seed mass; SSD, stem specific density.

From: Deep learning and citizen science enable automated plant trait predictions from photographs