A crucial part of the methods used today for wildlife monitoring is image processing. Since most wild creatures are most active at night, it might be difficult to view them without specialist equipment. The availability of heat sensing technology and technical advancements have made it possible for...
Mahdi and Mahmood (2022) used YOLOV5 with an edge computing infrastructure for a fire and non-fire classification task. They created a wildfire image data set from internet videos. They randomly initialised the weight within the range [0,1] instead of the pre-trained weights from COCO. A ...
Figure 1 presents the overall architecture of our approach. First, a video stream is sent into an object detection component, which is a YOLOv3 CNN. YOLOv3 is pre-trained on ImageNet and fine-tuned for detecting temperate fish species using a custom dataset. This component detects the presenc...
These characteristics imply that one-stage detectors are better suited for real-time species classification of large quantities of wildlife trade47. In the present study, object detection models, SSD using eight different CNNs as backbone networks, were assessed to classify turtles imported into Korea...
The classification of individual features of wood microscopic images by manual annotation is typically performed using an object-detection model that contains backbone networks, neck networks, detection heads or other components, such as YOLO (You Only Look Once) [26], SSD (Single Shot MultiBox Det...
WilDect-YOLO: An efficient and robust computer vision-based accurate object localization model for automated endangered wildlife detection 2023, Ecological Informatics Show abstract A federated learning system with enhanced feature extraction for human activity recognition 2021, Knowledge-Based Systems Show ab...
To establish the benchmark dataset, we evaluated the dataset using several DL models, including YOLOv7, YOLOv8 and Faster-RCNN, to locate and classify weeds in crops. The performance of the models was compared based on inference time and detection accuracy. YOLOv7 and its variant YOLOv7-...
To enable the automatic collection of data for the experiment, a simulated environment was created to simulate four classes of wildlife found in South Africa: buffalo, elephants, rhino and zebra. The network structure for the detector network selected was an adapted version of the tiny YOLOv3 ...
In recent years, deep learning has significantly reshaped numerous fields and applications, fundamentally altering how we tackle a variety of challenges. A
Review on methods used for wildlife species and individual identification. Eur J Wildl Res. 2022;68(1):3. H assan et al. Journal of Big Data (2024) 11:135 Page 27 of 29 3. Vranken E, Mounir M, Norton T. Sound-based monitoring of livestock. In: Zhang Q, editor. Encyclopedia...