Insect Detection and Classification on AgroPest-12 Dataset

GROUP: COMP9517-Arcadia

Introduction

This project presents the development and comparison of several traditional machine learning methods and deep learning models for pest detection and classification in the AgroPest-12 dataset. The aim is to measure their performance to analyze strengths, weaknesses, computational cost, and explainability.

Dataset

We use the AgroPest-12 dataset, containing 12 pest classes with annotated bounding boxes.

• Source：https://www.kaggle.com/datasets/rupankarmajumdar/crop-pests-dataset

Team Contributions

Chenyi Li:

• YOLO - v11
  • Source：https://github.com/ultralytics/ultralytics
• Transformer - RT-DETR
  • Source：https://github.com/lyuwenyu/RT-DETR
• XAI for yolov11 & RT-DETR
  • Source for yolov11 XAI：https://github.com/rigvedrs/YOLO-V12-CAM
  • Source for rtdetr XAI：https://github.com/jacobgil/pytorch-grad-cam

Yiyang Shen:

• InceptionV3
• EfficentNetB0

Yutong Zhang:

• CNN - VGG
• CNN - GoogleNet
• CNN - ResNet
• CNN - DenseNet
• CNN - Faster R-CNN

Yian Zhu:

• YOLOv8 + MobileNet

Ruiying Ren:

• SIFT/LBP
• SVM/KNN

Environment Setup

pip install ultralytics grad-cam opencv-python ttach
pip install kagglehub pyyaml torch torchvision matplotlib opencv-python tqdm scikit-image numpy scikit-learn seaborn

References

@software{yolo11_ultralytics, author = {Glenn Jocher and Jing Qiu}, title = {Ultralytics YOLO11}, version = {11.0.0}, year = {2024}, url = {https://github.com/ultralytics/ultralytics}, orcid = {0000-0001-5950-6979, 0000-0003-3783-7069}, license = {AGPL-3.0} }

@misc{lv2023detrs, title={DETRs Beat YOLOs on Real-time Object Detection}, author={Wenyu Lv and Shangliang Xu and Yian Zhao and Guanzhong Wang and Jinman Wei and Cheng Cui and Yuning Du and Qingqing Dang and Yi Liu}, year={2023}, eprint={2304.08069}, archivePrefix={arXiv}, primaryClass={cs.CV} }

@misc{jacobgilpytorchcam, title={PyTorch library for CAM methods}, author={Jacob Gildenblat and contributors}, year={2021}, publisher={GitHub}, howpublished={\url{https://github.com/jacobgil/pytorch-grad-cam}}, }

The XAI for YOLOv11 of this project makes use of the YOLO-V12-CAM repository: https://github.com/rigvedrs/YOLO-V12-CAM