In an exclusive interview with Higher Education Review, Dr. M.M. Ramya, Dean at Agurchand Manmull Jain College shares her thoughts on Artificial Intelligence and Machine Vision integration in educational institutions, future education with AI and Machine Vision, and the impacts of AI integration. She is a professor, focusing on e-Search about soft computing, with particular emphasis on image processing applications, and has 15 publications in academic journals and conference proceedings.

How can AI-driven tools enhance personalized learning experiences for students of different learning styles?

Artificial intelligence and machine vision are defined as two different technologies. AI concentrates on enabling machines to learn from data, while machine vision is concerned with the processing and understanding of visual information like images. AI and machine vision play a vital role in providing personalized learning experiences in education. It has been difficult for faculty members to fulfill the various learning needs of each student in outsize classes. Each student in a class may have different learning capabilities, making it impossible for an individual faculty to modify lessons and tests for each student.

However, when AI and machine vision are combined, it's feasible to personalize the learning process and examinations for individual students. AI tools can analyze individual learning patterns.  Using AI, faculties can create customized tests and provide different learning experiences and strategies in the classroom that are adapted to the needs of each student.

How can we bridge educators’ lack of technical background gap to enhance educators’ ability to teach AI and machine vision advanced topics?

Students nowadays are increasingly showing interest in visual and activity-based learning approaches. To improve teaching and learning processes, especially teaching difficult concepts, model assessment, and multi-model teachings such as project-based learning, interactive exercises, and real-life activities engage students more effectively than traditional lecture-based teaching. One of the key challenges with this approach is providing timely and personalized feedback to each student, especially during interactive or outsize class activities.

When a faculty is unable to give individual attention to each student, live feedback and constructive feedback can be given in real-time with an automated tool. By incorporating these technologies, complex concepts can be broken down into multi-model teaching and multi-model assessments and make students more engaged in class.

How can AI models, which are only as good as the data they are trained on, avoid perpetuating or amplifying existing biases, especially in machine vision systems, to prevent inequitable outcomes for students?

The success of any AI model depends on the kind of data used for training and testing. AI will fail to give accurate results when the AI model is trained with less data, centralized data without exceptional cases, and different types of data. This applies to all domains, not only educational institutions. AI systems work efficiently when trained with efficient data. In education institutions, the solution is to move beyond, not solely depending on conventional AI algorithms which follow traditional methods of training and testing.

The Generative AI generates new knowledge without training on specific models and can learn based on new experiences. Addressing problems in academics includes a combination of multiple AI models without relying on a single AI model much like how teachers learn from daily experiences in the classroom.

How can educators meaningfully integrate AI and machine vision into existing curricula, ensuring these technologies align with learning objectives and complement traditional teaching methods without overwhelming or distracting students?

AI enhances personalized learning experiences in educational institutions. When integrating AI and machine vision tools, the advantage lies in offering more personalized learning experiences. This can lead to significant advancements in teaching, learning, and educational technology. However, the success of these tools relies on their thoughtful implementation. Unlike simpler automation tools where things are automated and reports are generated, such as those used for accounting, ERP systems, or developing databases, implementing AI in teaching and learning is more complex.

The challenge is offering personalized learning experiences. Also, a student with lower learning abilities cannot receive simple questions without compromising their value of education. AI and machine learning techniques will be utilized by teachers to teach while preserving the human aspects of education. Teachers will continue to play a crucial part in education whereas AI is an additional tool to improve teaching methods and techniques.

How can schools, especially those in underserved areas with budget constraints, afford AI and machine vision tools?

The implementation of AI and machine vision requires further hardware and technology which needs financial resources. Depending upon the institution's capability the AI tool and automation brought into institutions will vary. Every institution to pursue 100% automation or rely entirely on AI support is not necessary. The priority in all intuitions is to raise awareness about AI, its potential benefits, and its applications among all faculty members. AI is no longer limited to computing science programs; every educator, regardless of their field, should have a basic understanding of AI.

In addition, it is crucial to know about the requirements of AI and a system that needs to be automated through AI. Instead, ensure the evaluation of examination papers should be done by faculty members or AI. If an institution approaches AI only for classroom delivery or evaluating objective-type questions the implementation would stop there. AI Implementation costs are reduced according to the implementations in educational institutions. Each institution can adopt AI and machine vision in a phased manner by achieving an implementation progressively depending on AI awareness among faculties and students and based on the institution’s available financial resources.