Pre-trained Models (Crop AI)
From building AI models to continuously improving them
Pre-trained models are not just initial models.
They create value within a continuous improvement cycle that connects data collection, annotation, training, evaluation, and deployment.
Chloros provides the foundation for building and operating crop image analysis AI efficiently.
A Continuous AI Workflow from Data to Deployment
Pre-trained models deliver value throughout this entire cycle
Capture high-quality field imagery through drone flights
Efficiently create training data with AI-assisted auto-annotation using pre-trained models
Train and fine-tune models using your own field data
Validate accuracy and improve model quality through iterative refinement
Deploy analysis results to the field through visualization and practical application
Supported Models & Detection Examples
We provide pre-trained models for detecting crops, growth stages, and pest/disease damage across a range of crops and use cases.
Detection examples using pre-trained AI models:
Supported crops and detection targets are continuously expanding.
What Pre-trained Models Enable
The foundation for building and operating crop image analysis AI faster and more reliably
Start Faster
Start model development from a strong baseline instead of building from scratch. Reduce the burden of initial model development.
Improve with Data
Continuously improve and optimize models by incorporating new field data. Adapt models to your crops, field conditions, and evaluation targets.
Connect Development to Deployment
Go beyond model development. Connect training, evaluation, deployment, and visualization to build practical AI systems.
AI models are not built once and finished.
They evolve continuously with data.
This workflow is designed based on MLOps principles, while focusing on practical value for crop image analysis.
Start building crop AI with your own data
Use pre-trained models as a baseline and customize them for your crops, fields, and evaluation goals.