Advanced montane bird monitoring using self-supervised learning and transformer on passive acoustic data

Ecological Informatics, Год журнала: 2024, Номер unknown, С. 102927 - 102927

Опубликована: Ноя. 1, 2024

Язык: Английский

---

Use ResNet50V2 Deep Learning Model to Classify Five Animal Species

Jurnal JTIK (Jurnal Teknologi Informasi dan Komunikasi), Год журнала: 2023, Номер 7(4), С. 758 - 768

Опубликована: Дек. 30, 2023

This study employs the ResNet50V2 Deep Learning model for purpose of classifying five distinct animal species. To gain insights into model's proficiency in visual recognition, we conducted training and testing procedures on a dataset comprising diverse images The utilization this classification task is intended to discern distinctions among these species by leveraging distinctive characteristics present input images. A meticulous comprehensive procedure was undertaken model, employing fine-tuning techniques adjust its internal representation order accommodate characteristics. experimental findings illustrate capacity effectively categorize various with notable degree precision, thereby presenting encouraging outcomes potential broader contexts. emphasizes significant models, specifically ResNet50V2, comprehending identifying fauna through cues.

Язык: Английский

---

Using photographs and deep neural networks to understand flowering phenology and diversity in mountain meadows

bioRxiv (Cold Spring Harbor Laboratory), Год журнала: 2023, Номер unknown

Опубликована: Март 29, 2023

ABSTRACT Mountain meadows are an essential part of the alpine-subalpine ecosystem; they provide ecosystem services like pollination and home to diverse plant communities. Changes in climate affect meadow ecology on multiple levels, for example by altering growing season dynamics. Tracking effects change diversity through impacts individual species overall dynamics is critical conservation efforts. Here, we explore how combine crowd sourced camera images with machine learning quantify flowering richness across a range elevations alpine located Mt Rainier National Park, Washington, USA. We employed three techniques (Mask R-CNN, RetinaNet YOLOv5) detect wildflower taken during two seasons. demonstrate that deep can species, providing information photographed meadows. The results indicate higher just above tree line most which comparable patterns found using field studies. two-stage detector Mask R-CNN was more accurate than single-stage detectors YOLO, network performing best mean average precision (mAP) 0.67 followed (0.5) YOLO (0.4). methods anchor box variations multiples 16 led enhanced accuracy. also show detection possible even when pictures interspersed complex backgrounds not focus. differential rates depending abundance, additional challenges related similarity flower characteristics, labeling errors, occlusion issues. Despite these potential biases limitations capturing abundance location-specific quantification, accuracy notable considering complexity types picture angles this data set. therefore expect approach be used address many ecological questions benefit from automated detection, including studies phenology floral resources, complement wide approaches (e.g., observations, experiments, community science, etc.). In all, our study suggests metrics efficiently monitored combining easily accessible publicly curated datasets Flickr, iNaturalist).

Язык: Английский

---

Real-Time Sound Detection of Rose-Ringed Parakeet Using LSTM Network with MFCC and Mel Spectrogram

Опубликована: Ноя. 16, 2023

Nowadays, machine learning, deep and artificial intelligence are essential in identifying bird presence rice fields other agricultural applications. Automated identification techniques make it easier for farmers to monitor populations identify potential threats crops without human intervention. Using sound-based analysis, (AI)--driven systems can distinguish between species that benefit the environment those might cause harm. This research presents a unique approach realtime vocalization of Rose-Ringed Parakeets. A Long Short-Term Memory (LSTM) neural network is used proposed approach, improved by Mel Frequency Cepstral Coefficients (MFCC) Spectrogram features. The audio data was processed extract MFCC features, which were then utilized capture spectrum attributes shown parakeet sounds. LSTM framework Rose-ringed Parakeets from their auditory signals developed implemented using Raspberry Pi 4 B single-board computer. Because its adaptability, our method ideal real-world use situations like prompt detection control populations. Automating AI procedures reduces possibility error ensures continuous accurate monitoring system. Farmers researchers data-driven decisions enhance crop yields, preserve biodiversity, promote sustainable practices with these technologies.

Язык: Английский

---

Threshold of anthropogenic sound levels within protected landscapes in Kerala, India, for avian habitat quality and conservation

Scientific Reports, Год журнала: 2024, Номер 14(1)

Опубликована: Фев. 1, 2024

Abstract Anthrophony is an important determinant of habitat quality in the Anthropocene. Acoustic adaptation birds at lower levels anthrophony known. However, threshold anthrophony, beyond which biophony starts decreasing, less explored. Here, we present empirical results relationship between and four terrestrial soundscapes. The constancy predicted vector normalised anthropogenic power spectral density (~ 0.40 Watts/Hz) all study sites intriguing. We propose value as indicator avian acoustic tolerance level sites. findings pave way to determine permissible sound within protected landscapes directly contribute conservation planning.

Язык: Английский

---

BAT-CNN: BirdNet Assisted Training for CNN

Опубликована: Июнь 21, 2024

Язык: Английский

---

Acoustic Based Bird Species Classification Using Deep Learning

Опубликована: Июнь 28, 2024

Язык: Английский

---

Soundscape Design in an Urban Natural Park

Land, Год журнала: 2024, Номер 13(10), С. 1546 - 1546

Опубликована: Сен. 24, 2024

Urban natural parks represent a remarkable concept that evokes the coexistence of human habitation with wild environment, and associated interactions between territories. In this context, urban noise infringes upon soundscape, leading to various consequences for both realms. This study seeks characterize impact anthropic levels on biodiversity in Văcărești Park (Bucharest, Romania), utilizing on-site measurements software simulation techniques. The develop method evaluating integrative strategies mitigate traffic wildlife an park, without addressing specific effects perception communication individual species. By calibrating field laboratory results, more reliable data set will be used identify areas where biophonic environment is impacted by anthropogenic noise. Since human-generated park predominantly originates from road industrial sites, managing its propagation pathways could substantially improve park’s soundscape. Additionally, apply simulations reduction strategies, such as vegetation planting earthen embankments, obtain suitable solutions propose plausible effective actions authorities improving environment. research also serve basis long-term monitoring, allowing assessment evolution implemented measures over time.

Язык: Английский

---

Optimizing the frequency of question items for bird species in quiz-style online training

Ecological Informatics, Год журнала: 2024, Номер unknown, С. 102908 - 102908

Опубликована: Ноя. 1, 2024

Язык: Английский

---

Ecology & computer audition: Applications of audio technology to monitor organisms and environment

Heliyon, Год журнала: 2023, Номер 10(1), С. e23142 - e23142

Опубликована: Дек. 2, 2023

Among the 17 Sustainable Development Goals (SDGs) proposed within 2030 Agenda and adopted by all United Nations member states, 13th SDG is a call for action to combat climate change. Moreover, SDGs 14 15 claim protection conservation of life below water on land, respectively. In this work, we provide literature-founded overview application areas, in which computer audition – powerful but context so far hardly considered technology, combining audio signal processing machine intelligence employed monitor our ecosystem with potential identify ecologically critical processes or states. We distinguish between applications related organisms, such as species richness analysis plant health monitoring, environment, melting ice monitoring wildfire detection. This work positions relation alternative approaches discussing methodological strengths limitations, well ethical aspects. conclude an urgent research community greater involvement methodology future approaches.

Язык: Английский

---