Table 1 Summary of few-shot learning applications in agriculture.

From: A novel framework GRCornShot for corn disease detection using few shot learning with prototypical network

Ref.

Application

Key contributions

32

Plant and pest classification

Baseline with 36 comparison experiments using public datasets

33

Plant disease detection

90.12% accuracy using ViT in plant disease detection

34

Plant disease and pest detection

Achieved 96% accuracy; +6.74% severity estimation

38

Cow identification

99.5% mAP for cow location, 90.43% accuracy with 5 samples

39

Agricultural pest and disease detection

Reviewed few-shot applications; promising results in pest identification

40

Forest identification

0.62 IoU for 1-way 1-shot identification

41

Cotton growth state recognition

88% accuracy in 3-way 5-shot tasks

42

Cross-domain few-shot learning

Addressed domain gap; state-of-the-art performance in CDFSL benchmark

Back to article page