Leaf Counting
Organ Segmentation
Spike Detection
Dataset
We constructed a large-scale unlabeled wheat visual dataset, ImAg4Wheat, comprising 2,000 genotypes and 500 environmental experimental conditions, totaling 3 million images to capture the diversity in wheat phenotypes comprehensively.
State-of-the-art Performance
Frozen features produced by the models are also evaluated on various visual tasks, including six wheat-specific tasks (two at the canopy level and four at the organ level), along with two tasks on rice and two on other crops and weeds. Results are thoroughly compared against features generated from other non-wheat datasets using state-of-the-art models.
Contact
Shouyang Liu ([email protected]), Hao Lu ([email protected]), Yanfeng Ding ([email protected])