Advanced Neural Network Applications

YOLito automates mosquito detection and behavioral analysis, achieving high accuracy across species and assays.

BehaveAI detects and classifies objects and behaviors from motion using a color-from-motion encoding strategy, enabling efficient and accurate analysis with minimal annotation effort.

PlanktonFlow is a Python pipeline that simplifies deep-learning classification of plankton images, demonstrating that EfficientNet-B5 outperforms other CNNs and EcoTaxa in classification accuracy.

This study uses CNNs to classify 15 European fly families and quantify uncertainty, achieving 95.61% accuracy with EfficientNetB4.

This study develops a CNN-based system for mosquito species identification using wing images, achieving high accuracy and demonstrating practical usability in real-world conditions.

Fusing a small-grain expert detector, fine-tuning across microscopes, human-in-the-loop continual learning and zero-shot SAM-2 segmentation improve detection and isolation of fossil pollen on scanned slides.

DaphnAI uses deep learning to automate zooplankton identification and segmentation, achieving high accuracy and speed for ecological monitoring.

This paper develops a computer vision model to estimate the age of white-tailed deer using dental features from jawbone images, achieving 90.7% accuracy.

This paper develops a deep learning model for age classification of white-tailed deer using trail camera images, achieving higher accuracy than traditional methods and human experts.

This paper develops an AI pipeline using YOLOv8 and transfer learning CNNs to detect and classify coliform colonies on agar plates, achieving faster and more accurate results than traditional methods.

This study compares a computer vision model and human observers in detecting African mammals in camera trap images, finding the model reduces processing time by 82% but undercounts species compared to humans.

This paper uses deep learning to detect sexual dimorphism in sociable weavers, finding higher accuracy than humans and identifying beak regions as key for sex prediction.

The paper introduces a joint probabilistic model for human scene and object recognition using residual computation, showing functional specialization in coarse-to-fine processing.

This paper introduces MacqD, a deep learning model for detecting macaques in complex lab environments, achieving high F1-scores across various scenarios.

This study uses deep learning models to classify Drosophila melanogaster gender with high accuracy and integrates the models into a mobile app for real-time analysis.

AInsectID Version 1.1 is an insect species identification software using transfer learning with CNNs, achieving 99.65% validation accuracy.

This paper improves bumblebee species identification by integrating object detection with segmentation and classification, enhancing performance in complex images.

The paper develops a CNN for seashell classification to aid in ecosystem restoration by accurately identifying confiscated seashells and processing over 36,000 images.

This study improves dog breed classification accuracy using a combination of CNN features, PCA, GWO, and SVM, achieving 95.24% and 99.34% accuracy for 120 and 76 breeds respectively.

The paper investigates why deep networks rely on color and texture over shape for classification, showing that training methods bias networks toward simpler features.

This paper presents a deep learning pipeline to monitor arthropod abundance and floral cover using time-lapse camera data, showing a correlation between flower cover and arthropod activity.

This paper releases a dataset of 5.6 million arthropod images for training machine learning models in taxonomic identification.

The paper introduces a dataset and a PyTorch model for automated wildflower monitoring, achieving a high mAP score for species identification and counting.

This study compares YOLOv8 models trained on different preprocessing methods for dragonfly wing images to determine the most effective approach for species classification.

This paper improves YOLO's pollinator classification and detection on small datasets by grouping classes into broader categories, finding that 'Pollinator/Non-pollinator' grouping performs best.

SegColR automates image segmentation and color extraction using deep learning models, enabling efficient phenotypic data analysis from biodiversity images.

The paper develops a machine learning method to count Antarctic shag nests using RPAS imagery, achieving high accuracy with limited training data.

This paper develops WOLO, a deep learning model that estimates ant body mass from reference-free images, achieving accuracy comparable to humans and significantly faster.

The paper develops a mobile hybrid deep learning tool for classifying 3D Amazonian lizard species using a dataset from the Emilio Goeldi Museum, achieving high accuracy with MobileNetV3-Small and Linear SVM.

The paper demonstrates the use of transfer learning for snake classification, showing high accuracy in species identification using deep learning models.