Advanced Neural Network Applications

A pilot computer-vision model trained to identify slide-mounted scale insect pests from the family Rhizoecidae

The paper trains a computer-vision model to identify 16 species of root mealybugs using high-resolution images from museum specimens, achieving 79% F1 score.

DroneMegaDetector, A Generalised Model For Detecting Large Mammals In Aerial Imagery

The paper introduces a general-purpose object detector for large mammals in aerial drone imagery and evaluates its performance across diverse datasets and conditions.

WildObserve: Camera-trap software that learns species labels during review and estimates occupancy and abundance

WildObserve automates camera-trap image analysis, adapting to projects with minimal resources and improving species classification and occupancy estimates.

InsectDCT: A generalized pipeline for detection, taxonomic classification, and tracking of insects in camera-trap recordings

InsectDCT detects, classifies, and tracks insects in camera-trap footage using deep learning, achieving high accuracy in complex environments.

Advancing Wildlife Image Analysis: A Graph Attention Contrastive Learning Approach for Region-Specific Mammal Classification

This paper develops a region-specific AI model for mammal classification using graph attention contrastive learning, achieving high accuracy in recognizing local species and improving data cleaning efficiency.

FP8 Inference in Genomic Foundation Models: Theoretical vs. Realized Speedups on GenomeOcean

This study evaluates FP8 quantization in GenomeOcean, finding it preserves biological fidelity while improving efficiency, but throughput gains fall short of theoretical limits.

A wildlife behaviour detection model: a case study on the red-legged seriema

The paper presents a deep learning model for detecting red-legged seriema behavior in videos, achieving high accuracy with a small dataset and providing a reproducible workflow for ecologists.

PlumageParts: A fine-grained avian plumage segmentation dataset and benchmark for ecological image analysis

PlumageParts provides a dataset of 4,705 bird images annotated for nine plumage regions and benchmarks segmentation models, achieving 84.01% mean Intersection over Union with a frozen DINOv3 encoder.

Optimizing automated classification for zooplankton in coastal conditions: the impact of model selection, imaging instruments, and colour information

This study evaluates how model selection, imaging instruments, and color information affect zooplankton classification using CNNs, finding that EfficientNetV2S performs best and color information provides limited benefit.

EcoMorph: Universal morphological trait quantification from natural language prompts for ecological research

EcoMorph quantifies morphological traits from images using natural language prompts, achieving high accuracy across diverse ecological contexts and taxa.

Application of Machine Learning Tools for Waterbird Colony Monitoring Provides Gains in Precision and Temporal Efficiency

The paper compares four waterbird monitoring methods using drone imagery, finding that fully automated detection reduces annotation time by 98% while maintaining 87% accuracy.

Object-zoomed training of convolutional neural networks inspired by toddler development improves shape bias

The paper introduces object-zoomed training for CNNs to improve shape bias, aligning CNN performance with human-like object recognition strategies.

Integration of Deep-Learning and Species Distribution Models for Classification of Animal Species of the Brazilian Fauna

The paper integrates deep-learning models with species distribution models to improve animal classification in Brazilian fauna, achieving higher accuracy than standalone models.

LizardLens: A Two-Stage Deep Learning Pipeline for Detecting and Classifying Similar Species in Visually Complex Environments

LizardLens improves species identification in complex environments using a two-stage deep learning pipeline, achieving high accuracy for Anolis lizards.

TetraFuse: A Synergistic Four-Dimensional Dynamic Fusion Framework for Efficient and Robust Medical Image Classification

TetraFuse improves medical image classification by integrating four domains to balance accuracy and efficiency, achieving high performance with reduced computational demand.

Deep Blueprint: A Literature Review and Guide to Automated Image Classification for Ecologists

This paper provides a guide for ecologists to use deep learning for image classification, demonstrating a practical approach with open-source data and models.

InsectMorphoAI: A Deep Learning-Based Software for Automated Estimation of Insect Length, Volume, and Biomass

InsectMorphoAI automates insect length, volume, and biomass estimation from 2D images using two methods, achieving high accuracy and enabling efficient ecological research.

Towards a general Detector of terrestrial Arthropods in Natural backgrounds

The paper develops a robust arthropod detector for natural backgrounds, achieving high accuracy across diverse species and contexts.

Exposure to naturalistic occlusion promotes generalized, human-like robustness in deep neural networks

This study shows that training DNNs with natural occluders improves their robustness to real-world occlusion, unlike artificial occluders which do not replicate human-like performance.

Paying Attention to Other Animal Detections Improves Camera Trap Classification Models

This paper improves camera trap classification models by incorporating self-attention to leverage information from multiple animal detections, enhancing species classification accuracy in both mono- and multi-species scenarios.

Insect Size Matters: Using Image and Dimensions Together Improves Image Classification

This paper improves insect classification by integrating body size information into deep learning models, showing it enhances accuracy, especially in low-data and taxonomic misclassification scenarios.

A Lightweight, High-Throughput Classifier for North American Insects Using EfficientNet: Elytra 1.0

Elytra 1.0 is a lightweight, high-throughput insect classifier achieving high accuracy on North American insects using EfficientNet, suitable for edge deployment.

Automated AI image recognition tools improve the efficiency of aerial wildlife counts: A multi-species case study on breeding seabirds and pinnipeds at the sub-Antarctic Bounty Islands.

This study uses AI to automate aerial wildlife counts, achieving high accuracy and efficiency for seabird and pinniped populations in remote areas.

A General Method for Detection and Segmentation of Terrestrial Arthropods in Images

flatbug, a YOLOv8 instance-segmentation model with multi-scale tile-pyramid inference and polygon-based deduplication, segments terrestrial arthropods across 23 image subdatasets and transfers to held-out subdatasets with a modest F1 penalty.

Automated Specimen Triage for Dark Taxa: Deep Learning Enables Orientation, Sex Identification, and Anatomical Segmentation from Robotic Imaging

This paper uses deep learning to automate specimen triage by identifying sex, orientation, and anatomical regions from robotic imaging of Phoridae specimens.

A study on edge devices for image classification of the Tasmanian devil (Sarcophilus harrisii) for vaccine delivery

This paper evaluates edge devices for image classification of Tasmanian devils to optimize vaccine delivery systems.

MultiSpecies Canopy Segmentation: Interactive Machine-Learning and Pseudo-Labelling are key

This paper develops a method to train deep-learning models for multi-species canopy segmentation using FCNNs and pseudo-labeling, achieving high accuracy with reduced human effort.

Reducing Manual Labour in Forensic Microtrace Recognition with Deep Learning

This paper develops a deep learning model to automate microtrace recognition in forensic microscopy, reducing manual annotation needs by 65% through combined pretraining strategies.

Automated detection of macropods in Tasmania using drone surveys and convolutional neural networks (CNNs)

This paper develops a CNN-based model for detecting macropods in drone imagery, achieving high recall across diverse habitats and demonstrating the feasibility of deep learning for wildlife monitoring.

YOLito: A generalizable model for automated mosquito detection

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