metadata
license: apache-2.0
language:
- en
pretty_name: iPheno-120K
task_categories:
- image-classification
- object-detection
- visual-question-answering
tags:
- agriculture
- pest-disease-detection
- crop-ripeness
- crop-freshness
- crop-object-detection
size_categories:
- large
Dataset Description
iPheno-120K is a comprehensive agricultural image dataset derived from both publicly available sources and manually collected real-world samples. The dataset is characterized by high ecological authenticity and task diversity. Specifically, iPheno-120K covers four representative tasks: pest and disease detection, ripeness assessment, freshness evaluation, and crop object detection. It comprises nearly 40 major crop types with more than 70,000 original images, further augmented by a 120K Visual Question Answering (VQA) dataset generated for training based on these images.
Tasks:
- Pest and Disease Detection: Identifying pests and diseases affecting crops.
- Ripeness Assessment: Evaluating the ripeness of crops.
- Freshness Evaluation: Assessing the freshness of crops based on visual cues.
- Crop Object Detection: Detecting and identifying various crop objects in images.