Research Saturday Week 8: The Sesame Dataset

​This week at Research Saturday, we will review the Sesame Plant Segmentation Dataset, a specialized dataset developed for precise segmentation of sesame plants in agricultural fields.

​Unlike traditional detection approaches, this dataset uses segmentation to clearly outline the shape and structure of sesame plants. It is fully converted to the YOLO segmentation format, making it compatible with Ultralytics YOLOv8 to YOLO11 and other segmentation focused AI models.

​The dataset contains training and validation images captured across different growth stages, lighting conditions, and field perspectives. This supports model robustness and reduces background interference during training. We will explore the dataset design, labeling approach, structure, and potential areas for expansion.

​We will also discuss the range of applications this dataset enables, including crop monitoring, growth analysis, smart farming automation, and agricultural research.

• ​Date: Saturday, 29 November 2025

• ​Time: 8 PM WAT

• ​Dataset: Sesame Plant Segmentation Dataset - https://www.datum.africa/datasets/sesame-plant-segmentation-dataset

• ​Presenter: Sunusi Ibrahim Muahammad (CTO, Ejaztech AI)
  

​Join the conversation and explore how datasets like this can advance AI powered agriculture across Africa.