Can Face recognition tech identify you while wearing a face mask?

Recognizing faces while wearing a face mask is considered the most difficult facial occlusion challenge because it occludes a large area that usually covers around 60% of the frontal face which contains rich features including the nose and mouth. This is emerging problem which higlighted during the pandemic period. I was able to develop a appropriate solution to overcome the above issue through a research study. The developed occlusion invariant face recognition system is published as a open work to the public.

Rapid development and breakthrough of deep learning in the recent past have witnessed the most promising results from face recognition algorithms. But they fail to perform far from satisfactory levels in the unconstrained environment during the challenges such as varying lighting conditions, low resolution, facial expressions, pose variation and occlusions (Wen et al., 2016). Among the challenges, Occlusion is one of the most challenging problems faced by current face recognition algorithms. Moreover, Face mask makes higher inter-class similarity and inter-class variations due to covering a large area of the face which tricks the facial verification. process of face recognition systems (Guo and Zhang, 2019). Ongoing Face Recognition Vendor Test (FRVT) (Ngan, Grother and Hanaoka, 2020) is a performance evaluation report based on 89 face verification algorithms while wearing face masks (Ngan, Grother and Hanaoka, 2020). The study (Damer et al., 2020), concludes their results on the effect of wearing a face mask which given strong points of performance degradation on face recognition systems, showing the necessity of the development of mask capable face recognition systems.

Facial recognition has been studied for several years and it’s a longitudinal & preliminary task in computer vision (Best-Rowden and Jain, 2018). Among the biometrics, methodologies face recognition has a better edge on the accuracy, efficiency, usability, security, and privacy to recognize the identity non-intrusively. Furthermore, Facial recognition is widely preferred over potential candidates because of its contactless authentication behaviour which helps to keep good hygiene practices.(Zeng, Veldhuis and Spreeuwers, 2020). Face recognition is widely used in personal device log-on, passport checking, ATMs, border control, forensics, law enforcement, surveillance, and national security (Ejaz et al., 2019).

The study, by (Damer et al., 2020) concludestheir evaluationsidentified strong signs of an adverse effect of face masks on existing face recognition algorithms and occlusion invariant face recognition algorithms, pointing out the failure of them to identify faces while wearing a mask and the necessity of developing appropriate face recognition solutions. Furthermore, Ongoing Face Recognition Vendor Test (FRVT) is a performance evaluation report based on 89 face recognition algorithms on masked faces. If the mask covers around 70% of the occluded area most algorithm’s failure rate is increased by 20% to 50% according to the National Institute of Standards and Technology, United States (Ngan, Grother and Hanaoka, 2020).

The recent surveys on the effect of face mask occlusion on face recognition precisely illustrate that existing face recognition systems and occlusion invariant face recognition systems are failed to handle face mask occlusions. However, most of the work done in small occlusions such as spectacles, sunglasses and partial captures commonly appears in in-the-wild capture conditions. In the existing research literature, there are few efforts are focused to handle face mask occlusions that are not met expected accuracy, efficiency, usability and security. This research is focused on building an effective novel facial recognition algorithm with the capability of recognizing faces while wearing a face mask in real-time to overtake barriers of occlusion invariant face recognition with growing incentives of wearing face masks without reducing well-known qualities of facial recognition.

Adaptive learning app for learning java. Gamification and recommendation are key to provide adaptive learning experience to users.

It's a Smart Learning App for Java including Performance monitor and Recommendation System based on Student performance. Developed the recommendation system using sentimental analysis and Content-based Filtering. The Python(Flask) API catered the services to the react native mobile application. Flask Api is developed using RESTful concepts and hosted on AWS. NLP workflow is designed using NumPy, pandas, sci-kit-learn and nltk Machine learning libraries. React native frontend integrated back end to establish a connection to an online database using SQLAlchemy. Python, Python Flask, AWS, and Git were the technologies and tools used.

An adaptive learning system aims at providing instruction tailored to the current status of a learner, differing from the traditional classroom experience. The latest advances in technology make adaptive learning possible, which has the potential to provide students with high-quality learning benefit at a low cost. A key component of an adaptive learning system is a recommendation system, which recommends the next material (video lectures, practices, and so on, on different skills) to the learner, based on the psychometric assessment results and possibly other individual characteristics. An important question then follows: How should recommendations be made? To answer this question, a mathematical framework is proposed that characterizes the recommendation process as a Markov decision problem, for which decisions are made based on the current knowledge of the learner and that of the learning materials. In particular, two plain vanilla systems are introduced, for which the optimal recommendation at each stage can be obtained analytically.

This Project as carried out through requirement gathering phase through surveys.The software Requirement specification is presented with the project.

Library Management System is also known as an integrated library system that manages all the activities such as acquisition, circulation, stock verification, cataloging, etc. Library Management System Software helps the students to remain up-to-date with the library database and other resources. Digital Library System is an Advance Library Software that helps in fast and effortless gathering of all information in the library. As we all know the library is an important part of all schools and universities and the demand for Library Automation Software in increased with the advancement of the system. Make your library a smart library with Library Administration Software which will turn your library into a digital system where you can easily access the database. As the digital revolution has already taken a boom, every school and college also wants to get advance and remain up-to-date with the current market demand. The library system has taken gear due to Government educational policies and systems. The government provides funding to the schools and colleges for the installation of the Library Automation System.

Features of the Digital Library Management System are: Easy way to enter book data. Full catalogue, circulation, and acquisition system for library stock management. Web-based OPAC (Online Public Access Catalogue) system. Simple, clear search interface for all users. Multilingual and multi-user support. Export and import records. A modern integrated library management system. Print own barcodes

Concurrent programming application for shared asynchronous printing system including real time refilling written using java concurrency concepts.

The system has following set of features.

• Threads vs processes
• Multithreading vs Multitasking
• synchronus vs asynchronus processing
• Threads Creation
• Thread Lifecycle
• Thread Synchronization
• Deadlocks
• Java Monitor
• Thread Pool
• Java Semaphore

Let's see the beauty of pathfinding.

This is a 2D maze path solver visualizer implemented with Java.

Mobile web application fully facilitated with the capabilities of the train ticket booking system and more exciting features.

We designed our eRAIL, train mobile web application fully facilitated with the capabilities of the train ticket booking system and more exciting features. It is based on popular and historical traveling destinations in Sri lanka. We used Gamification to improve the passenger’s knowledge about the destination through playing a game. Our application has more additional features to attract the potential customers to encourage visiting these destinations and inspire other potential passengers by challenging them through the App. In our application we smoothly handle user experience principles to attract users towards the app. The following works are Carried through the project.

• Style Guide
• UI Flow Diagram
• Iphone UI Designs
• Ipad UI Designs
• Prototype Demo
• Implementation