AGE AND GENDER BASED MOVIE RECOMMENDATION SYSTEM USING FACIAL RECOGNITION

AGE AND GENDER BASED MOVIE RECOMMENDATION SYSTEM USING FACIAL RECOGNITION

ABSTRACT:-

            Recommendation systems help users to find and select items e.g., books, movies, restaurants from the huge number of things available in front of us. Movie recommendation systems additionally have an equivalent problem since people have completely different expectations whereas observation of a movie and recommendation isn’t potential based mostly on the annotations given by the different users. A movie recommendation system provides a level of comfort and personalization that helps the user move higher with the system and watch movies that cater to his desires.

            One of the most known methods is collaborative filtering so notice the movies those users have viewed and given a good rating that has not been viewed by specific users, recommending movies to users has nice profit, but applying chance for specific movie based mostly on genre is done, that can increase accuracy in recommendation engine. This recommendation system is based on age, emotion and gender recognition. For this recognition CNN(convolutional neural network) algorithm is used. Collaborative Filtering is used for customer segmentation. In this system movie is recommended to users based on his previous history. If a user watched a comedy movie, this system recommends the user all comedy movies. This system recommends movies according to your age. This system recommends cartoons movie to children and meaningful movies for younger mens. If a user likes to watch bollywood’s movie system recommends all bollywood movies to the user.

            This is where the recommender system comes into image wherever the content providers recommend users the content consistent with the users’ emotion. During this paper we’ve planned a movie recommender system MovieMender. The objective of MovieMender is to supply correct movie recommendations to users. Sometimes the essential recommender systems think about one of the subsequent factors for generating recommendations; the preference of users (i.e content based mostly filtering) or the preference collaborative users (i.e cooperative filtering). to create a stable and correct recommender system a hybrid of content based mostly filtering also as collaborative filtering is used.

OBJECTIVES OF THE PROJECT:-

The main objectives of this system are:

• Gender, emotion and age recognition is used for customer segmentation. Gender, emotion and age is done based on image analysis.
• We add that classification to the already available database for customer segmentation.  After customer segmentation, we recommend movies according to segments and develop chatbot related to that.
• Gender, emotion and age recognition with the help of image analysis and classification Customer segmentation using image input, login input and/or database.
• Movie recommendation system based on the segment favorites, and “one who saw this also saw” method. 

EXISTING SYSTEM:-

The recommendation system has been the standard for analysis over the past decade, so many business models have been built on it, several standard learning models and artificial intelligence have been used to study the exhibition recommendation program as a K-Mean and soft collection High Resolution used for visual presentation from user ratings, as Root-Means-Square Error (RMSE) has been used for outcome analysis, in paper collaborative sorting is often a given recommendation, a well-known algorithmic rule read more Ten years of exhibition recommendation program, these strategies same user with the same interest to make it easier to suggest another film the same user has seen and given a reasonable rating, that show isn’t It does not come from a ray current nzisi. On paper 2 types are considered for users to rate search for the same users or user tags that have access to information or use the show name to retrieve movie information from the IMDB server.

PROPOSED SYSTEM:-

              Nowadays, the development of recommendation programs will be a major test site that attracts many scientists and researchers around the world. Promotional systems have been used in many places as well as track, movies, books, details, search queries, and marketing products. Collaborative filtering algorithm is one of the most popular self-paced strategies for rs, wishing to search users carefully as the strongest to elevate objects. Collaborative filtering (cf) with a few alternating squares (als) of rules that are the most important techniques used to build a movie recommendation engine. A set of rules is one of the matrix models that solves a linked cf which is taken into account because the values are within the user’s matrix object list. As there is a need to perform an analysis on the als algorithm by selecting fully specified parameters that may ultimately make it easier to build a raw movie recommendation engine. At some point in this paper, we tend to advise on a movie-sponsored movie recommendation machine. This analysis makes special selections of the parameters of als algorithms to influence the performance of the rs solid structure. From the results, drawing conclusions in line with the selection of als algorithms parameters can also influence the overall performance of the film recommendation engine. Version analysis ended with aberrations of very unfamiliar metrics such as execution time, root means rmse errors of the standard prediction, and the degree to which the most effective translation has evolved.

Gender, emotion and age recognition is used for customer segmentation. Gender, emotion and age is done based on image analysis. We add that classification to the already available database for customer segmentation.  After customer segmentation, we recommend movies according to segments and develop chatbot related to that. It takes both image and login inputs. It not only caters to the interests of the person who has logged in but to the actual person sitting in front of the computer or smart television. It also has a chatbot feature to answer any and all questions that the user might have.

MODULES:

• User Registration: User has to register to watch movies that cater to his needs.
• User Login: User can login to system to watch movies that cater to his needs.
• Facial recognition: With the help of CNN algorithm and face landmark user face will be captured by camera and feature extraction of face will be done which will result in obtaining output such as age, gender and emotion.
• Movie Recommendation: Based on customer segment collaborative filtering recommend some movies to the customer.
• Tflearn model generation: Based on question-answer dataset model get trained with convolutional neural network.
• Chatbot: Chatbot will solve customer queries which is trained by using tflearn algorithm.

ADVANTAGES:-

• Conclusion is drawn according to the selection of parameters of ALS algorithms which can affect the performance of building a movie recommender engine.
• Experiment results show significant improvement in scalability and performance over the most efficient existing solutions for item recommendation.
• One resolution can be to recommend movies similar to the genre of top rated movies. If the new movie falls in that genre, it will get discovered. However, in this resolution, we will need to build a system based on the genre of the movies.
• We give a detailed design and development process, and test the stability and high efficiency of the experiment system through professional tests.

HARDWARE AND SOFTWARE REQUIREMENTS

HARDWARE:-

• Processor: Intel Core i3 or more.
• RAM: 4GB or more.
• Hard disk: 250 GB or more.

SOFTWARE:-

• Operating System : Windows 10, 7, 8.
•  
•  
• Spyder, Jupyter notebook, Flask.
•