Coming soon the StomataHubDetector (SHD); a web-based platform for stomata detection, trained using the entire stomata dataset available on StomataHub, comprised of over 20 species and highly varied images. SHD v1 will make available a combination of common deep learning architectures including, but not limited to, YOLO, VGG and R-CNN for bounding box detection.
SHD v1 aims to provide access to the core bounding box architectures to encourage future research to focus on more informative methods for stomata analysis and make the need for additional bounding box methods redundant. Instead future work should look towards methods such as semantic or instance segmentation are one particular example which can be used to obtain more in-depth traits