Animal Recognition System Based on Convolutional Neural Network

Trnovszký, Tibor; Kamencay, Patrik; Orjesek, Richard; Benčo, Miroslav; Sýkora, Peter

doi:10.15598/aeee.v15i3.2202

Cited by 55 publications

(30 citation statements)

References 4 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…The detection / recognition of objects stays a difficult challenge and there is no specific approach that can address all problems in a stable and efficient way. Broadly speaking, animal detection algorithms is used as a binary classification task [17] for animal detection. Then the classifier will decide if the sample is the animal, when a new input image is given.…”

Section: Animal Detection Methods In Field Of Image Processingmentioning

confidence: 99%

“…Compares the animal and confirms or continues to refuse the identity of the discovered animal (the one -to-one matching). Even though verification and identification often share the same classification algorithms, the two modes are intended for specific applications [17 ]. ⮚ Identification-compares the picture of the animal with all other animals in the database and gives the matches a categorized list (one to n matching).…”

mentioning

confidence: 99%

See 1 more Smart Citation

Scrutiny of Methods for Image Detection and Recognition of Different Species of Animals

2019

IJRTE

View full text Add to dashboard Cite

Animal detection-based study is useful in many real-life applications. Techniques involved in animal detection are useful in observing the locomotive behavior of the engaged animal and in result it prevent harmful interruption of animals in residential areas. There are some branches of research in animal detection. Some of these branches will therefore be discussed in this journal. Humans have developed many algorithms and techniques to gain a better understanding of animal behaviour. Therefore, for early preventive measures, these technologies can also serve as a warning system for humans from encroachment of dangerous wild animals. Such tasks can be reduced to three main branches, namely animal detection, tracking and recognition. Through these papers, new approaches for study and a variety of technologies/algorithms implemented in the past are identified and appropriate ways for solving the research gaps are suggested to fill the gap.

show abstract

Section: Animal Detection Methods In Field Of Image Processingmentioning

confidence: 99%

mentioning

confidence: 99%

Scrutiny of Methods for Image Detection and Recognition of Different Species of Animals

2019

IJRTE

View full text Add to dashboard Cite

show abstract

“…With the advent of Deep Learning and the improved image analysis performance associated with convolutional neural networks (CNNs), robust and accurate analysis of ecological images is now possible [81][82][83]. However, large diverse training sets are required, and the use of computer vision may have been limited thus far by the significant amount of time and money needed to create these trainings sets.…”

Section: Integrating Ai Into Camera Trap and Citizen Science Work Flowsmentioning

confidence: 99%

Innovations in Camera Trapping Technology and Approaches: The Integration of Citizen Science and Artificial Intelligence

Green

Rees

Stephens

et al. 2020

Animals

View full text Add to dashboard Cite

Simple Summary: Camera traps, also known as "game cameras" or "trail cameras", have increasingly been used in wildlife research over the last 20 years. Although early units were bulky and the set-up was complicated, modern camera traps are compact, integrated units able to collect vast digital datasets. Some of the challenges now facing researchers include the time required to view, classify, and sort all of the footage collected, as well as the logistics of establishing and maintaining camera trap sampling arrays across wide geographic areas. One solution to this problem is to enlist or recruit the public for help as 'citizen scientists' collecting and processing data. Artificial Intelligence (AI) is also being used to identify animals in digital photos and video; however, this process is relatively new, and machine-based classifications are not yet fully reliable. By combining citizen science with AI, it should be possible to improve efficiency and increase classification accuracy, while simultaneously maintaining and promoting the benefits associated with public engagement with, and awareness of, wildlife.Abstract: Camera trapping has become an increasingly reliable and mainstream tool for surveying a diversity of wildlife species. Concurrent with this has been an increasing effort to involve the wider public in the research process, in an approach known as 'citizen science'. To date, millions of people have contributed to research across a wide variety of disciplines as a result. Although their value for public engagement was recognised early on, camera traps were initially ill-suited for citizen science. As camera trap technology has evolved, cameras have become more user-friendly and the enormous quantities of data they now collect has led researchers to seek assistance in classifying footage. This has now made camera trap research a prime candidate for citizen science, as reflected by the large number of camera trap projects now integrating public participation. Researchers are also turning to Artificial Intelligence (AI) to assist with classification of footage. Although this rapidly-advancing field is already proving a useful tool, accuracy is variable and AI does not provide the social and engagement benefits associated with citizen science approaches. We propose, as a solution, more efforts to combine citizen science with AI to improve classification accuracy and efficiency while maintaining public involvement.

show abstract

“…En [19] se comparan métodos de clasificación basados en rasgos, contra el enfoque de jerarquía de rasgos (aprendizaje profundo, DL); en ese sentido, en este artículo se reconocen especies animales usando métodos basados en rasgos, tales como: Análisis Lineal de Discriminantes, Análisis de Componentes Principales, Histogramas de Patrones Binarios Locales y Máquinas de Vectores de Soporte (LDA, PCA, LBPH y SVM, respectivamente, todos derivados de sus siglas en inglés) [20].…”

Section: Trabajo Relacionadounclassified

Identification of Rodent Species Using Deep Learning

Seijas

Montilla²,

Frassato³

2019

CyS

View full text Add to dashboard Cite

En este artículo se describe un sistema identificador de especies de roedores usando herramientas computacionales del paradigma de aprendizaje profundo. Las especies identificadas son 4 tipos diferentes de roedores y la identificación se logra usando técnicas de inteligencia artificial aplicadas a imágenes de estos roedores en su hábitat natural. Estas imágenes fueron captadas, mediante sistemas de cámaras activadas en modo automático, ocultas en el hábitat natural de las especies en estudio, en condiciones, tanto, de luz del día como también nocturnas y etiquetadas por expertos. El conjunto de imágenes acopiada constituye el conjunto de datos para entrenamiento de tipo supervisado, de 1411 imágenes de 4 clases. El identificador desarrollado, es un clasificador multiclase, basado en el paradigma de aprendizaje profundo del amplio tema del aprendizaje automático, lo que permite construir un sistema de altísimo desempeño. El clasificador consta de tres etapas conectadas en cascada, siendo la primera etapa, una etapa de pre-procesamiento, luego, está una red neuronal convolucional (CNN, de sus siglas en inglés) para extracción de rasgos, implementada con una arquitectura pre-entrenada usando el método de aprendizaje por transferencia; específicamente, la CNN usada es la conocida VGG-16; a esta segunda etapa, se le conecta como etapa siguiente y final, una máquina de vectores de soporte (SVM, de sus siglas en inglés) que actúa como la etapa clasificadora. A fines de comparación, los resultados se contrastan contra modelos de identificación automáticos anteriormente publicados, los resultados logrados con nuestro identificador son significativamente superiores a los alcanzados en investigaciones previas en el tema. Palabras claves.Identificación de especies, aprendizaje profundo, red neuronal convolucional preentrenada.Abstract. In this article, we describe a rodent species identification system using computational tools of the deep learning paradigm. The identified species are 4 different types of rodents and the identification is achieved using artificial intelligence techniques applied to images of these rodents in their natural habitat. These images were captured, using camera systems activated in automatic mode, hidden in the natural habitat of the species under study, under both daylight and nighttime conditions and labeled by experts. The collected image set constitutes the data set for supervised training of 1411 images of 4 classes. The identifier developed is a multiclass classifier, based on the deep learning paradigm of the broad topic of machine learning, which allows to build a high performance system. The classifier consists of three stages connected in cascade, being the first stage, a pre-processing stage, then, there is a convolutional neural network (CNN) for feature extraction, implemented with a pre-trained architecture using the method of learning by transfer; specifically, the CNN used is the well-known VGG-16; to this second stage, a support vector machine (SVM) is connected as the next and fin...

show abstract

Animal Recognition System Based on Convolutional Neural Network

Cited by 55 publications

References 4 publications

Scrutiny of Methods for Image Detection and Recognition of Different Species of Animals

Scrutiny of Methods for Image Detection and Recognition of Different Species of Animals

Innovations in Camera Trapping Technology and Approaches: The Integration of Citizen Science and Artificial Intelligence

Identification of Rodent Species Using Deep Learning

Contact Info

Product

Resources

About