Login/Register
ISSN 2583-2913
Home
About Us
Editorial Board
Current Issue
Archives
In Press
Submit Manuscript
Instructions
AWSHR
Contact Us
Latest Updates
Articles are invited for the first issue of Indian Journal of Science and Research.
Join Indian Journal of Science and Research as reviewer, editorial board member.
Congratulations!!! To all the members for Joining editorial board of Indian Journal of Science and Research.
Why Publish With Us
Fast reviewing process: Adopting the fast reviewing process we communicate with authors at all stages till publication. Proactive teams of reviewers from National and International front help us to complete the process within the time.
Article In Press
Title:
COMPUTATIONAL APPROACHES FOR PLANT DISEASE DETECTION: A COMPARATIVE STUDY USING IMAGE ANALYSIS AND DEEP LEARNING
Author:
Win Htet Soe, Min Zayya, Min Kaung Khant, Mya Thidar Aung and Hlaing Htake Khaung Tin*
Keyword:
Plant diseases, RGB, CNN, SVM, GLCM, sustainable agriculture.
Page No:
173-183
DOI:
https://doi.org/10.5281/zenodo.21756792
Abstract:
The occurrence of plant diseases is the critical issue for food security in the modern world. Traditional approaches to detecting diseases, including visual inspection, microscopy and culture-based te
sts, are accurate yet slow and complex. Nevertheless, the progress in computation techniques, specifically image processing and machine learning algorithms, allows automating and accelerating the approach to disease detection. The following work provides a comparison of different computational techniques aimed at detecting plant diseases using 1,000 images of leaves from tomato and potato plants that suffer from blight, leaf spot, and early blight. The feature extraction methods include the use of RGB normalization and GLCM technique, combined with classical machine learning (Support Vector Machine and Random Forest models) and neural network (CNNs) approaches. The experiment shows that CNNs demonstrate higher performance than classical machine learning methods, as their accuracy reaches 94, precision equals to 92, recall 93, and F1-score 92.5, while SVM and RF models show the accuracy only 85-88. It means that computational techniques can be used as a part of an IoT/mobile application to detect plant diseases in real time and manage them according to sustainable farming needs.
Download PDF
Title:
COMPARATIVE ANALYSIS OF LISTENER PERCEPTION AND EMOTIONAL RESPONSE TO AI-GENERATED AND HUMAN-COMPOSED MUSIC ACROSS MULTIPLE GENRES
Author:
Hlaing Htake Khaung Tin*, Aye Chan Myae, Thin Yanant and April Thet Su
Keyword:
AI-Generated Music, Human-Composed Music, Listener Perception, Emotional Response, Music Genre Analysis
Page No:
184-191
DOI:
https://doi.org/10.5281/zenodo. 21757046
Abstract:
Recent developments in artificial intelligence have affected the creative industry, such as music composition, and present questions about the role of human creativity and listener perceptions. This r
esearch offers an empirical comparison of audience perceptions and emotional experiences of AI-composed versus human-composed music across various genres such as classical, rock and pop. Using a survey approach, a representative sample of listeners was asked to rate a set of audio clips, without knowledge of their authorship. Participants' assessments were captured through Likert-scale surveys that addressed aspects of listener perception including emotional response, creativity, authenticity and general listening quality. Data were explored handling descriptive and inferential statistical techniques to compare perceptions between types of music and genres. The results show that although AI-generated music is technically proficient and engaging, human-composed music consistently scores better in terms of emotional engagement and authenticity. Further analysis also revealed genre-specific differences with AI-generated music scoring better in structured genres rather than more free-form styles. The findings of this analyze will be instrumental in guiding musicians, technologists, and scholars on how to apply AI in their musical creations.
Download PDF
Title:
A REVIEW OF MULTIBAND IMAGING TECHNIQUES IN AGRICULTURE
Author:
Poe Ei Phyu, Phyu Phyu Htun and Hlaing Htake Khaung Tin
Keyword:
Multiband Imaging, Agriculture Technology, Precision Agriculture, Multispectral Sensors
Page No:
194-203
DOI:
https://doi.org/10.5281/zenodo.21757275
Abstract:
The multispectral imaging method entails collecting and processing an image through different wavelength bands on the electromagnetic spectrum. The imaging method is entirely different from the conven
tional imaging technique since it does not objective one specialized wavelength band or channel such as visible light. In its place, it collects and processes various spectral bands of information either simultaneously or one by one. This provides a deeper insight into the subject matter being analyzed. The multi-band or multi-spectral imaging process involves gathering various bands of wavelengths to form a full spectrum of the view of an object or image. In this case, each wavelength corresponds to a specified region on the electromagnetic band, including visible light, infrared light, and ultraviolet light. This means that multiband imaging provides a more comprehensive analysis of the situation than single-channel images that have a limited scope of wavelengths. It is vital in numerous areas, particularly those involving the use of machine vision cameras such as agriculture, medicine, and others. Due to population increases, natural resource exhaustion can be expected; therefore, farming production must be improved.
Download PDF
