A CNN Based Model for Venomous and Non-venomous Snake Classification [Elektronisk resurs]
-
Progga, Nagifa Ilma (författare)
-
1st International Conference on Applied Intelligence and Informatics (AII 2021), Nottingham, UK (online), July 30-31, 2021
-
Rezoana, Noortaz (författare)
-
Hossain, Mohammad Shahadat (författare)
-
Islam, Raihan Ul, 1981- (författare)
-
Andersson, Karl, 1970- (författare)
-
Luleå tekniska universitet Institutionen för system- och rymdteknik (utgivare)
- Publicerad: Springer, 2021
- Engelska.
-
Serie: Communications in Computer and Information Science, 1865-0929 1865-0929
-
Ingår i: Applied Intelligence and Informatics. ; 216-231
-
Läs hela texten
-
Läs hela texten
-
Läs hela texten
- Relaterad länk:
-
http://www.ltu.se/ (Värdpublikation)
Sammanfattning
Ämnesord
Stäng
- Snakes are curved, limbless, warm blooded reptiles of the phylum serpents. Any characteristics, including head form, body shape, physical appearance, texture of skin and eye structure, might be used to individually identify nonvenomous and venomous snakes, that are not usual among non-experts peoples. A standard machine learning methodology has also been used to create an automated categorization of species of snake dependent upon the photograph, in which the characteristics must be manually adjusted. As a result, a Deep convolutional neural network has been proposed in this paper to classify snakes into two categories: venomous and non-venomous. A set of data of 1766 snake pictures is used to implement seven Neural network with our proposed model. The amount of photographs even has been increased by utilizing various image enhancement techniques. Ultimately, the transfer learning methodology is utilized to boost the identification process accuracy even more. Five-fold cross-validating for SGD optimizer shows that the proposed model is capable of classifying the snake images with a high accuracy of 91.30%. Without Cross validation the model shows 90.50% accuracy.
Ämnesord
- Natural Sciences (hsv)
- Computer and Information Sciences (hsv)
- Computer Sciences (hsv)
- Naturvetenskap (hsv)
- Data- och informationsvetenskap (hsv)
- Datavetenskap (datalogi) (hsv)
- Pervasive Mobile Computing (ltu)
- Distribuerade datorsystem (ltu)
Genre
- government publication (marcgt)
Indexterm och SAB-rubrik
- Snake
- CNN
- Data augmentation
- Deep learning
- Transfer learning
- Cross validation
Inställningar
Hjälp
Uppgift om bibliotek saknas i LIBRIS
Kontakta ditt bibliotek, eller sök utanför LIBRIS. Se högermenyn.