Project title: An Enhanced CNN-VAE-based DL Framework for Sea Turtle Identification
Position: PhD Candidate
Benefits:
100% tuition fee waiver
RM2,850 monthly stipend (up to 42 months)
Supervisors: Dr Veera Ragavan S. , Dr. M Ayoub Juman, Dr. Vishnu Monn
Project Description:
Malaysia, renowned for its rich biological diversity, is home to various endangered species, including sea turtles. Baby turtles upon hatching embark on a remarkable journey, navigating vast oceans before returning to their birthplace, a process known as 'the lost years.' Traditional tracking methods using physical tags have proven invasive, unreliable, and impossible to attach at birth. However, anatomical research established that sea turtles possess unique identifier patterns - akin to human fingerprints - such as the numbers and pattern of plates on their shells and the scale pattern on their heads, which enables visual recognition. This project aims to comprehensively understand the morphological changes in sea turtle scale patterns, explicitly focusing on the critical early stages, leading to the development of an Enhanced Convolutional Neural Network-Variational Auto Encoder (CNN-VAE)--based Deep Learning (DL) Framework. First, morphological changes in sea turtle scale patterns and their correlation with age will be investigated. Next, a cutting-edge CNN-VAE-based DL Framework, capable of projecting a baby turtle's image through its various developmental stages, will be developed, resulting in an extensive database for the age classification of wild turtles. Additionally, this initiative seeks to establish a comprehensive Citizen Science (CS) platform, harnessing crowd-sourced environmental data to enhance the database for reverse classification, fostering a collaborative and inclusive effort in gathering vital information for sea turtle conservation. Benefits include accurate and efficient age classification while minimizing the need for manual or invasive tracking methods, thus contributing to the conservation of these endangered species. This novel and innovative approach will aid global turtle recognition, facilitating conservation efforts and aligning with Malaysia's conservation efforts, supporting local initiatives and contributing to achieving UN Sustainable Development Goals 14-Life below Water.
Requirements:
Open to Malaysian or International Candidates
Hold an excellent Bachelor degree with CGPA of at least 3.67 (First class) or equivalent in Electrical / Electronics Engineering, Mechanical Engineering, Mechatronics Engineering, Computer Engineering, or similar disciplines.
Students who are expected to graduate in the near future are encouraged to apply.
Having prior journal publications is preferable
Strong analytical skills and mathematical background.
Fundamental knowledge in the areas of AI / IoT
Strong skills in programming/ using tools, such as Python
Excellent, motivated, and self-driven and Hands-on.
Please contact Dr Veera Ragavan S. via email : veera.ragavan@monash.edu