Related Experiment Video
Updated: Jun 14, 2025

05:57
Long-term Video Tracking of Cohoused Aquatic Animals: A Case Study of the Daily Locomotor Activity of the Norway Lobster Nephrops norvegicus
Published on: April 8, 2019
6.8K
GooseDetectlion: A Fully Annotated Dataset for Lion-head Goose Detection in Smart Farms
Yuhong Feng1, Wen Li2, Yuhang Guo1
1The College of Computer Science and Software Engineering, Shenzhen University, Shenzhen, 518060, China.
Scientific Data
|September 7, 2024
Summary
This study introduces GooseDetectlion, the first large dataset for AI in smart farming, featuring 2,660 images of lion-head geese. Object detection models, particularly YOLO v5, show promising results for reducing farmer workload.
Area of Science:
- Computer Science
- Agricultural Science
- Artificial Intelligence
Background:
- Developing Artificial Intelligence (AI) models for smart farming requires extensive datasets to automate tasks and reduce manual labor.
- Existing animal datasets may lack the scale and annotation density needed for robust AI model training in agricultural contexts.
Purpose of the Study:
- To introduce GooseDetectlion, a novel and comprehensive dataset for lion-head goose detection.
- To establish a challenging benchmark for evaluating object detection algorithms in an agricultural setting.
- To facilitate the development of AI-powered smart farming solutions.
Main Methods:
- Collection of 2,660 images from 6 cameras deployed on a goose farm.
- Annotation of 98,111 bounding boxes by 50 volunteers.
- Evaluation of six state-of-the-art object detectors, including YOLO v5, on the GooseDetectlion dataset.
Main Results:
- GooseDetectlion offers higher capacity and density compared to existing animal datasets.
- You Only Look Once version 5 (YOLO v5) demonstrated superior performance in detection precision, model size, and inference efficiency.
- The dataset provides a challenging benchmark for mainstream object detectors.
Conclusions:
- The GooseDetectlion dataset is a valuable resource for advancing AI in smart goose farming.
- YOLO v5 is a highly effective model for goose detection, contributing to reduced farmer workload.
- This work supports the development of more efficient and automated agricultural practices.

