Livestock Fish Larvae Counting using DETR and YOLO based Deep Networks
0
Sign in to get full access
Overview
- This paper presents a deep learning-based approach for counting livestock and fish larvae.
- The researchers explore the use of DETR (Detect Everywhere Transform) and YOLO (You Only Look Once) models for this task.
- The goal is to develop an efficient and accurate system for automated livestock and fish larvae counting.
Plain English Explanation
The paper focuses on using deep learning models to count the number of livestock and fish larvae in images or videos. Deep learning is a type of artificial intelligence that can learn to recognize patterns in data, like the appearance of animals or organisms in images.
The researchers tested two specific deep learning models: DETR and YOLO. DETR is a model that can detect objects of interest, like livestock or fish larvae, in an image. YOLO is another type of object detection model that can quickly identify and count the number of objects present.
The key idea is to use these deep learning models to automate the process of counting livestock and fish larvae. This could be useful for farmers, fisheries, and researchers who need to monitor the populations of these animals. By automating the counting process, it can save time and provide more accurate data compared to manual counting methods.
Technical Explanation
The paper explores the use of two deep learning architectures for livestock and fish larvae counting:
-
DETR (Detect Everywhere Transform): This is a transformer-based object detection model that can identify and locate objects of interest in an image. The researchers adapted the DETR model to work with livestock and fish larvae data.
-
YOLO (You Only Look Once): YOLO is a popular real-time object detection system that can quickly identify and count the number of objects in an image. The researchers used different versions of the YOLO model, including YOLOv5 and YOLOv8, to detect and count livestock and fish larvae.
The paper compares the performance of the DETR and YOLO models on several datasets, including images of livestock and fish larvae. The researchers evaluated the models' accuracy in detecting and counting the animals, as well as their inference speed and computational efficiency.
Critical Analysis
The paper presents a promising approach for automating the counting of livestock and fish larvae using deep learning models. However, there are a few potential limitations and areas for further research:
-
Dataset size and diversity: The performance of deep learning models is heavily dependent on the quality and size of the training data. The paper does not provide detailed information on the size and diversity of the datasets used, which could impact the models' generalization capabilities.
-
Real-world deployment: While the models showed good performance in the controlled experimental setting, their performance in real-world scenarios with varying lighting, occlusions, and other environmental factors may need further investigation.
-
Generalization to other species: The paper focuses on livestock and fish larvae, but it would be interesting to see how the models perform on a wider range of animal species that are of interest to farmers, fisheries, and researchers.
-
Computational efficiency: While the YOLO models are known for their speed, the computational requirements of the DETR model may need to be optimized for real-time or on-edge deployment in resource-constrained environments.
Conclusion
This paper presents a promising approach for automating the counting of livestock and fish larvae using deep learning models. The researchers explored the use of DETR and YOLO architectures, which showed good performance in detecting and counting the animals.
The results of this work could have significant implications for farmers, fisheries, and researchers who need to monitor the populations of these animals. By automating the counting process, the deep learning models can save time and provide more accurate data compared to manual counting methods.
While the paper demonstrates the potential of this approach, further research is needed to address the limitations and explore the generalization of the models to other species and real-world scenarios.
This summary was produced with help from an AI and may contain inaccuracies - check out the links to read the original source documents!
Related Papers
0
Livestock Fish Larvae Counting using DETR and YOLO based Deep Networks
Daniel Ortega de Carvalho, Luiz Felipe Teodoro Monteiro, Fernanda Marques Bazilio, Gabriel Toshio Hirokawa Higa, Hemerson Pistori
Counting fish larvae is an important, yet demanding and time consuming, task in aquaculture. In order to address this problem, in this work, we evaluate four neural network architectures, including convolutional neural networks and transformers, in different sizes, in the task of fish larvae counting. For the evaluation, we present a new annotated image dataset with less data collection requirements than preceding works, with images of spotted sorubim and dourado larvae. By using image tiling techniques, we achieve a MAPE of 4.46% ($pm 4.70$) with an extra large real time detection transformer, and 4.71% ($pm 4.98$) with a medium-sized YOLOv8.
Read more8/12/2024
0
A method for detecting dead fish on large water surfaces based on improved YOLOv10
Qingbin Tian, Yukang Huo, Mingyuan Yao, Haihua Wang
Dead fish frequently appear on the water surface due to various factors. If not promptly detected and removed, these dead fish can cause significant issues such as water quality deterioration, ecosystem damage, and disease transmission. Consequently, it is imperative to develop rapid and effective detection methods to mitigate these challenges. Conventional methods for detecting dead fish are often constrained by manpower and time limitations, struggling to effectively manage the intricacies of aquatic environments. This paper proposes an end-to-end detection model built upon an enhanced YOLOv10 framework, designed specifically to swiftly and precisely detect deceased fish across extensive water surfaces.Key enhancements include: (1) Replacing YOLOv10's backbone network with FasterNet to reduce model complexity while maintaining high detection accuracy; (2) Improving feature fusion in the Neck section through enhanced connectivity methods and replacing the original C2f module with CSPStage modules; (3) Adding a compact target detection head to enhance the detection performance of smaller objects. Experimental results demonstrate significant improvements in P(precision), R(recall), and AP(average precision) compared to the baseline model YOLOv10n. Furthermore, our model outperforms other models in the YOLO series by significantly reducing model size and parameter count, while sustaining high inference speed and achieving optimal AP performance. The model facilitates rapid and accurate detection of dead fish in large-scale aquaculture systems. Finally, through ablation experiments, we systematically analyze and assess the contribution of each model component to the overall system performance.
Read more9/4/2024
🤿
0
FishNet: Deep Neural Networks for Low-Cost Fish Stock Estimation
Moseli Mots'oehli, Anton Nikolaev, Wawan B. IGede, John Lynham, Peter J. Mous, Peter Sadowski
Fish stock assessment often involves manual fish counting by taxonomy specialists, which is both time-consuming and costly. We propose FishNet, an automated computer vision system for both taxonomic classification and fish size estimation from images captured with a low-cost digital camera. The system first performs object detection and segmentation using a Mask R-CNN to identify individual fish from images containing multiple fish, possibly consisting of different species. Then each fish species is classified and the length is predicted using separate machine learning models. To develop the model, we use a dataset of 300,000 hand-labeled images containing 1.2M fish of 163 different species and ranging in length from 10cm to 250cm, with additional annotations and quality control methods used to curate high-quality training data. On held-out test data sets, our system achieves a 92% intersection over union on the fish segmentation task, a 89% top-1 classification accuracy on single fish species classification, and a 2.3cm mean absolute error on the fish length estimation task.
Read more7/1/2024
0
Multi-Species Object Detection in Drone Imagery for Population Monitoring of Endangered Animals
Sowmya Sankaran
Animal populations worldwide are rapidly declining, and a technology that can accurately count endangered species could be vital for monitoring population changes over several years. This research focused on fine-tuning object detection models for drone images to create accurate counts of animal species. Hundreds of images taken using a drone and large, openly available drone-image datasets were used to fine-tune machine learning models with the baseline YOLOv8 architecture. We trained 30 different models, with the largest having 43.7 million parameters and 365 layers, and used hyperparameter tuning and data augmentation techniques to improve accuracy. While the state-of-the-art YOLOv8 baseline had only 0.7% accuracy on a dataset of safari animals, our models had 95% accuracy on the same dataset. Finally, we deployed the models on the Jetson Orin Nano for demonstration of low-power real-time species detection for easy inference on drones.
Read more7/2/2024