A Case Study of Testing an Image Recognition

Chuanqi Tao, Chuanqi Tao
2021 Journal of Visual Language and Computing  
A B S T R A C T High-quality Artificial intelligence (AI) software in different domains, like image recognition, has been widely emerged in people's daily life. They are built on machine learning models to implement intelligent features. However, the current research on image recognition software rarely discusses test questions, clear quality requirements, and evaluation methods. The quality of image recognition applications becomes more and more prominent. A three-dimensional(3D)
more » ... decision table can help users to conduct classification-based test requirement analysis and modeling for any given mobile apps powered with AI functions in detection, classification, and prediction. This paper presents a case study of a realistic image recognition application called Calorie Mama using manual testing and automation testing with a 3D decision table. The study results indicate the proposed method is feasible and effective in quality evaluation. © 2021 KSI Research time application of software, such as some software that testing the software is very important to verify that the is used to help with surgeries in the hospital. Therefore, the product meets requirements and specifications. Software testing ensures the correctness, integrity, and high quality of the software by checking errors or bugs and fixing them in the initial design. This paper focused on testing an image recognition application called Calorie Mama utilizing both manual testing and automation testing. Calorie Mama is a smartphone app that runs on Android and IOS devices. It uses deep learning to recognize food from food images and track nutrition based on the food in the image. It calculates the calorie based on that. We evaluated the performance, correctness, and quality of the app using both manual testing and automation testing. This paper is written to provide our perspective views on image recognition software testing and quality evaluation. The paper is organized as follows. Section 2 discusses the review of AI software testing and image recognition. The third part elaborates methodologies of manual and automation testing. Then, the fourth part shows a case study of testing Calorie Mama APP using these two methods and presents the comparative results of test efficiency and coverage. At last, section 5 gives the conclusion.
doi:10.18293/jvlc2021-n2-194 fatcat:e7opbrma2ngbllbncfnbrdidii