Green Pearl Education and Management Corporation (GPEMC) has recently sanctioned a significant consultancy project to SRM University-AP, titled “Multilingual Minutes of the Meeting Generation.” This project aims to develop innovative solutions for generating multilingual minutes of meetings, ensuring effective communication, and understanding across language barriers.

The Consultancy Project has been assigned to Dr. Ashu Abdul and Dr. Dinesh Reddy Vemula, Assistant Professors for a period of 7 months from the Department of Computer Science and Engineering (CSE). Their expertise in language processing and computational linguistics will play a crucial role in the successful execution of this project.

Joining them in this endeavour is Mr Phanindra Kumar, a dedicated PhD scholar, and a team of talented B.Tech. students, including Mr Shivansh Goel, Mr Md Ahmad Raza Khan, Ms Subrabala Dash, Mr Nithish Sri Ram, Mr Hadi Mahmood, and Mr Phanidra Kumar. Their combined skills and enthusiasm will contribute to the development of innovative solutions for multilingual meeting documentation.

The Consultancy Project has been sanctioned with an outlay of Rs 48,91,786/- as consultancy charges for a period of 7 months. This funding will support the research and development efforts required to deliver efficient and accurate multilingual minutes of meetings.

The objective of this project is to create an automated system that can generate minutes of meetings in multiple languages, ensuring accurate and comprehensive documentation. By overcoming language barriers, this project aims to facilitate effective communication and collaboration in diverse organizational settings.

The collaboration between SRM University-AP and GPEMC highlights the commitment of both institutions to research and innovation. This consultancy project presents an exciting opportunity for the faculty members and students to contribute to real-world problem-solving and make a substantial impact in the field of language processing.

Abstract

The project ‘Multilingual Minutes of the Meeting Generation’ aims to develop an innovative application that automates the creation of meeting minutes, action plans, and summaries in multiple languages. Leveraging advanced research in natural language processing and speech recognition technologies, our application seeks to streamline the tedious and time-consuming task of minute-taking. It will do so by accurately identifying, associating, and documenting statements with their respective speakers during meetings using speech diarization techniques.

The primary objective is to enhance the efficiency and accuracy of meeting minutes while supporting diverse linguistic needs and eliminating human error, thereby catering to global business environments. The system will improve the meeting outcomes for all participants by identifying the action items from the meeting. This project intends to deliver an effective solution for enterprises by combining the latest technological advances with a user-centric design.

Explanation of the Research in Layperson’s Terms

The project called “Multilingual Minutes of the Meeting Generation” is about creating a new tool that automatically writes down what happens in meetings—like the minutes, action items, and summaries—and can do this in several languages. This tool uses the latest technology in understanding and processing human speech, which helps make the task of recording meeting details much faster and easier.

The special feature of this tool is its ability to figure out who is speaking and accurately attach their words to them in the meeting notes. This makes sure the meeting records are both correct and detailed. The main goal here is to make creating and keeping track of meeting records more efficient and accurate, help people who speak different languages, and reduce mistakes that can happen when humans take down notes.

This system is especially useful in today’s global business setting, where people from different parts of the world need to work together smoothly. It aims to lessen the work involved in manually writing meeting notes and make the results easier for everyone to use and understand. Overall, the project promises to blend cutting-edge technology with designs that focus on the needs of the users to provide a practical tool for businesses.

Practical Implementation or the Social Implications Associated with the Research

The practical implementation of this project is particularly beneficial for enterprises, institutions, and professional settings where meetings are frequent. By automating minute-taking, it allows participants to focus more on discussion and decision-making, rather than note-taking. This leads to more productive meetings and ensures that important details are not lost or misinterpreted. Ultimately, this technology enhances communication and documentation accuracy across diverse linguistic and cultural landscapes in business and professional environments.

Collaborations

This project is being developed in collaboration with the members of the Next Tech Lab AP at SRM University-AP

Future Research Plans

The future plans for this project involve integrating the application with widely used platforms such as Zoom, Google Meet, and others that support video/audio conferencing. This integration aims to seamlessly incorporate the automated minute-taking functionality into these platforms, enhancing accessibility and usability for users across various companies and institutions. By expanding compatibility with popular conferencing tools, the project seeks to further streamline meeting documentation processes and cater to a broader user base in diverse professional settings.

A research paper titled “Detection of Diabetic Retinopathy (DR) severity from fundus photographs: An Ensemble Approach using weighted average”, has been published by Dr Mahesh Kumar Morampudi, Assistant Professor of Computer Science and Engineering, in the Arabian Journal for Science and Engineering.

Diabetic Retinopathy is a common diabetic disease that affects the retina and can result in blindness if not treated initially. Deep learning (DL) based models are proposed to detect the blood abnormalities in the retinal tissue due to diabetes mellitus obtained from fundus camera. The drawback of these models is the lack of performance. To address this, we propose to automate the process of detection of the severity of Diabetic Retinopathy (DR) using ensembles of pre-trained models, thus exploring the power of transfer learning in the field of automated diagnosis. Deep learning models perform well when the model is trained on a large amount of data. In this regard, we also put forth data augmentation and preprocessing techniques to generate synthetic images and to improve image quality. Extensive experimental results on publicly available databases illustrate that the proposed ensemble model achieves fair accuracy when compared to existing models. Thus, the proposed model shows good scope for deployment in real-time diagnosis.

Every year multiple people are diagnosed with diabetes. Diabetes is a chronic disease that affects several organs of the human body namely the eyes, kidneys, heart etc. Diabetic Retinopathy (DR) is a situation induced by diabetes in which severe loss happens to the retinal blood vessels that can ultimately lead to blindness. Regular diabetic retinopathy screening is hence needed for detecting it in advance. In the present situation, a trained clinician or an ophthalmologist is required to identify diabetic retinopathy (DR) by the existence of lesions related to the vascular abnormalities induced by the disease. The ophthalmologist needs to evaluate and examine digital colour fundus images of the retina. So, it is a very tedious and sluggish process. He/She needs more time to diagnose DR. The diagnosis of the disease by any manual means seems to be tiresome and usually results in errors. To overcome this limitation, we propose a model to automate the process of detection of the severity of DR using ensembles of pre-trained models, thus exploring the power of transfer learning in the field of automated diagnosis.

The research group believes that the study helps ophthalmologists to identify diabetic retinopathy at its early stage accurately, as a result, the chance of losing the vision due to diabetic retinopathy can be reduced. The work is done in collaboration with Dr Mulagala Sandhya, Assistant professor, NIT-Warangal. In the future, Dr Mahesh Kumar plans to work on a project related to Privacy-preserving Biometric Authentication.

The technological advancements in the medical domain have aided in the effective collection of data such as personal information, clinical history, and disease symptoms of patients. However, the accumulation of massive quantity of data may cause errors in the diagnosis of the disease. A chronic disease dataset may be comprised of numerous symptoms and attributes where not all of them are of equal importance in disease diagnosis. A few of those attributes may be less relevant or redundant.

Through her research, Dr Priyanka from the Department of Computer Science and Engineering proposes metaheuristics driven attribute optimization techniques that can be implemented in optimizing chronic disease datasets to achieve optimal efficiency in disease risk prediction which can help in proper medical diagnosis. Her paper titled “A Decisive Metaheuristics Attribute Selector Enabled Combined Unsupervised-Supervised Model for Chronic Disease Risks Assessment” was published in the Q1 journal Computational Intelligence and Neuroscience.

This research can be used to develop a decision support system to assist medical experts in the effective analysis of chronic diseases in a cost-effective manner. The system model may be used to assist medical experts in the efficient diagnosis of chronic disease risks. In future, the research study can be further enhanced to validate the model on more complex heterogeneous datasets with varying sizes and structures. Also, deep learning methods can be employed using image-based real-time datasets.

Abstract of the Research

Advanced predictive analytics coupled with an effective attribute selection method plays a pivotal role in precise assessment of chronic disorders risks in patients. In this paper, a novel buffer enabled heuristic a Memory based Metaheuristics Attribute Selection (MMAS) model is proposed, which performs local neighbourhood search for optimizing chronic disorders data. Heart disease, breast cancer, diabetes and hepatitis are the datasets used in the research. Upon implementation of the model, a mean accuracy of 94.5% using MMAS was recorded and it dropped to 93.5% if clustering was not used. The average precision, recall and f-score metric computed were 96.05%, 94.07% and 95.06% respectively. The model also has a least latency of 0.8 sec. Thus it is demonstrated that chronic disease diagnosis can be significantly improved by heuristics based attribute selection coupled with clustering followed by classification.

Figure 1: The proposed Metaheuristics attribute selector based classification model for chronic disorders detection

Figure 2: Accuracy analysis of MMAS method on different disease datasets