Skip to content

SHRISH01/Arthropod-Taxonomy-Orders-Object-Detection

Repository files navigation

Arthropod Taxonomy Orders Dataset

About the Dataset

Dataset Overview

Species identification from an image is a complex problem. Image classification assumes there is only one species in the image. However, the goal is to identify all species present in an image. Thankfully, biologists and taxonomists have systematically classified and ordered organisms in a taxonomic hierarchy.

The ArTaxOr dataset focuses on arthropods, which include insects, spiders, crustaceans, centipedes, millipedes, etc. With over 1.3 million described species of arthropods, creating a single dataset to cover them all is not feasible. Instead, the identification problem is broken down into multiple tasks:

Content

The dataset consists of images of arthropods in jpeg format and object boundary boxes in json format. There are between one and 50 objects per image.

This dataset is actively maintained, and new orders will be added on a regular basis. Currently, the following orders are covered with at least 2000 objects per order:

Araneae (spiders), adults, juveniles

Coleoptera (beetles), adults

Diptera (true flies, including mosquitoes, midges, crane file etc.), adults

Hemiptera (true bugs, including aphids, cicadas, planthoppers, shield bugs etc.), adults and nymphs

Hymenoptera (ants, bees, wasps), adults

Lepidoptera (butterflies, moths), adults

Odonata (dragonflies, damselflies), adults

Dataset Description

Dataset Description


Annotation

I have annotated the data using JSON files to provide precise boundary boxes for the objects in each image.

Annotation Testing

Testing of annotations was performed to ensure accuracy and consistency.

Annotation Testing


Model Training

The model was trained using YOLOv8 on an NVIDIA Tesla T4 GPU for efficient performance.

Model Training


Prediction and Testing

For prediction and testing, Flask was used to deploy and test the trained model.

Prediction - Step 1

Initial prediction result:

Prediction Step 1

Prediction - Final Result

Final refined prediction result:

Prediction Final Result


Pipeline

  1. Dataset Annotation
  2. Model Training using YOLOv8
  3. Deployment using Flask
  4. Prediction and Testing