No abstract

No abstract

Music has become an integral part of our day-to-day life. With the advancement in technology, lots of businesses are nowadays using recommendation systems for their benefit like in e-commerce, selling books, movies, and video recommendations. This helps businesses to get monetary benefits and the users to get better services. One of the best ways to suggest songs would be according to the people’s mood. In this project, we aim to interpret the user’s mood at a particular time using what they speak and then suggest songs based on the analysis done by the system. The best thing about music is that nothing can be more relaxing than a pleasant melody. Due to all these things, we choose to do this Song Recommendation system according to the mood of the user.

The traffic flow forecasting is very important aspect of traffic predication and congestion. It alleviates the increasing congestion problems that cause drivers to shorten the travelling duration required and prevent financial loss. Increasing congestion is one of the severe problems in big city areas. The aspect of traffic prediction is that it may give drivers to plan their traveling time and traveling path, based on the predictive data information they have. The aim is to design locally weighted regression model by proposing a method, which is a combination of Genetic algorithm and locally weighted regression method. This model helps to achieve optimal prediction performance under various traffic condition parameters. The time series model is used to predict the forecast value for the accurate assumption of the traffic volume generation according to the road capacity. The GA model results show these kind of predictions always be useful for highway road authorities.

The transition of technology over the years increases its application in new domains such as face detection using a pre-trained model. The cosmetic industry is most popular now with the need for automating the experience of a trial virtually. The objective of this project is to create an artificial intelligence-enabled makeup tool which provides the same level of experience. This is done using semantic separation in the profile preview using a pre-trained model called BiSeNet. This model identifies facial features into its classes. The makeup tool changes the color of lips and hair giving a realistic look for a better realisation. This provides a virtual experience to try on different choices of the user without the need to visit a store.