Skip to content

ethanclewis/Lettuce-NPK

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

14 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

FNNPK

Lettuce NPK Dataset image files.

  • Fully Nutritional (FN): 12 images
  • Nitrogen Deficient (-N): 58 images
  • Phosphorus Deficient (-P): 66 images
  • Potassium Deficient (-K): 72 images

Fine-Tuning Iterations

Indivudal notebooks detailing training loop and model parameters, as well as results, for each progressive, fine-tuning iteration in between the Initial Model and the Final Model.
A summary of all relevant parameters are listed at the top of each notebook.
All models trained on Nvidia GeForce GTX 1660 Ti GPU.

Initial Model

Initial Model.ipynb

Notebook detailing training loop and model parameters for initial model training attempt, as well as results.
Useful for direct comparison to the Final Model.
Trained on Apple M1 Chip.

mpk_model_v3.pth

Saved model.
The Initial Model was trained on Apple M1 Chip, rather than Nvidia GeForce GTX 1660 Ti GPU, hence the alternative saved model path.

EDA

Notebook provides basic visualization of the Lettuce NPK Dataset and its class distribution.
The Transformation Playground details our process of deriving an appropriate image augmentation pipeline.

Final Model.ipynb

Notebook detailing training loop and model parameters for final model training attempt, as well as results.
Useful for direct comparison to the Initial Model.
Trained on Nvidia GeForce GTX 1660 Ti GPU.

About

Repo details the process of leveraging the open-source "Lettuce NPK Dataset" to develop an image classification model capable of identifying various nutrient deficiencies in lettuce plants. The model is customized for optimal performance within the OpenSTEM environment: a portable hydroponic gardening unit.

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors