Spatial Transformer Network YOLO Model for Agricultural Object Detection

Yash Zambre,Ekdev Rajkitkul,Akshatha Mohan,Joshua Peeples

2024-09-16

Abstract:Object detection plays a crucial role in the field of computer vision by autonomously locating and identifying objects of interest. The You Only Look Once (YOLO) model is an effective single-shot detector. However, YOLO faces challenges in cluttered or partially occluded scenes and can struggle with small, low-contrast objects. We propose a new method that integrates spatial transformer networks (STNs) into YOLO to improve performance. The proposed STN-YOLO aims to enhance the model's effectiveness by focusing on important areas of the image and improving the spatial invariance of the model before the detection process. Our proposed method improved object detection performance both qualitatively and quantitatively. We explore the impact of different localization networks within the STN module as well as the robustness of the model across different spatial transformations. We apply the STN-YOLO on benchmark datasets for Agricultural object detection as well as a new dataset from a state-of-the-art plant phenotyping greenhouse facility. Our code and dataset are publicly available.

Computer Vision and Pattern Recognition,Artificial Intelligence,Machine Learning

What problem does this paper attempt to address?

The problem that this paper attempts to solve is the performance challenges of the YOLO model in handling complex scenes, partially occluded or low - contrast objects in agricultural object detection. Specifically, the paper proposes a new method, that is, integrating Spatial Transformer Networks (STNs) into the YOLO model to improve the performance of the model. Through this method, the STN - YOLO model aims to enhance the focus on important regions of the image and improve the spatial invariance of the model before the detection process, thereby improving the performance of object detection both qualitatively and quantitatively. The paper pays special attention to object detection in the agricultural field, such as applications in crop disease detection, pest detection and crop harvesting. In these application scenarios, the YOLO model has some limitations due to various spatial transformations. Therefore, by introducing STN to improve the robustness of the model to spatial transformations, the accuracy and efficiency of agricultural object detection are further improved. In addition, the paper also introduces a new high - quality Plant Growth and Phenotyping (PGP) dataset. This dataset contains multispectral images with features such as different heights, lighting conditions, plant sizes and shape changes, which further increases the challenges of agricultural object detection. Through experiments on this new dataset, the effectiveness of the STN - YOLO model is verified.

Spatial Transformer Network YOLO Model for Agricultural Object Detection

An Object Detection Method Based on Improved YOLOX

A Convolution with Transformer Attention Module Integrating Local and Global Features for Object Detection in Remote Sensing Based on YOLOv8n

An improved YOLOv7 model based on Swin Transformer and Trident Pyramid Networks for accurate tomato detection

HR-YOLOv8: A Crop Growth Status Object Detection Method Based on YOLOv8

Swin-Transformer-Based YOLOv5 for Small-Object Detection in Remote Sensing Images

C3TB-YOLOv5: integrated YOLOv5 with transformer for object detection in high-resolution remote sensing images

YOLO-DCTI: Small Object Detection in Remote Sensing Base on Contextual Transformer Enhancement

Cross stage partial connections based weighted Bi-directional feature pyramid and enhanced spatial transformation network for robust object detection

R-YOLO: A YOLO-Based Method for Arbitrary-Oriented Target Detection in High-Resolution Remote Sensing Images

Tomato Anomalies Detection in Greenhouse Scenarios Based on YOLO-Dense

Real-time and effective detection of agricultural pest using an improved YOLOv5 network

STC-YOLO: Small Object Detection Network for Traffic Signs in Complex Environments

Object Detection Based on Swin Deformable Transformer-BiPAFPN-YOLOX

Spatial-Transformer and Cross-Scale Fusion Network (STCS-Net) for Small Object Detection in Remote Sensing Images

TPH-YOLOv5++: Boosting Object Detection on Drone-Captured Scenarios with Cross-Layer Asymmetric Transformer

TPH-YOLOv5: Improved YOLOv5 Based on Transformer Prediction Head for Object Detection on Drone-captured Scenarios

Remote sensing object detection based on a combination of a CNN and the Swin transformer

Adaptive Dual-Stream Sparse Transformer Network for Salient Object Detection in Optical Remote Sensing Images

YOLO-TLA: An Efficient and Lightweight Small Object Detection Model based on YOLOv5

A fast accurate fine-grain object detection model based on YOLOv4 deep neural network