Resources

Datasets & Code

Open-source tools, datasets, and code repositories powering reproducible research across agriculture and autonomous vehicles.

Datasets

Open Datasets

Annotated imagery spanning livestock, plant phenotyping, and quality control — freely available for research use.

Pig posture keypoint detection

Swine Posture Dataset

3,000+ training images with 23,000+ annotated instances for multi-view pig posture recognition. Active Kaggle challenge with cash prizes.

Kaggle Challenge → Overhead piglet tracking with BoT-SORT bounding box annotations in farrowing crate

Gilt Activity Dataset

Multi-modal RGB and RGB-D dataset of gilt (female breeding pig) activities for precision livestock farming.

HuggingFace → Dry bean quality samples

Dry Bean Dataset

Bean color and quality analysis imagery (777 MB) for Bean Split Ratio classification and variety identification.

MSU / CVPPP → Dense leaf segmentation

Dense-Leaves Dataset

Dense foliage segmentation dataset (827 MB) with individual leaf annotations for plant phenotyping research.

MSU →

Multi-modal Plant Phenotyping

Imagery database spanning RGB, depth, and fluorescence modalities for plant phenotyping and photosynthesis analysis.

MSU CVLab →

Pig Body Condition Scoring

Ultrasound-based body condition scoring tool for swine. Paper, code, and dataset coming soon.

Annotation Tool →

Code

Open-Source Repositories

Implementation code for our most-cited methods, released for the research community.

CLOCs — Camera-LiDAR Fusion

Reference implementation for the CLOCs late-fusion 3D object-detection pipeline. Most-cited work from the lab (696+ citations).

GitHub →

Depth Coefficients

Depth completion implementation from CVPR 2019 — color-image-guided upsampling of sparse LiDAR depth.

GitHub →

Python Intro (Teaching)

Teaching repository used for ECE/CSE 434 Autonomous Vehicles — Python refresher, OpenCV, PyTorch, ROS basics.

GitHub →