ISSN:1005-3026

VIDANALYTICS: ML-BASED VIDEO ANALYSIS FOR IDENTIFICATION OF TARGET AUDIENCE IN BUSINESS DEVELOPMENT

Dhruv Mehta

Dept. of Computer Engineering Dwarkadas. J. Sanghvi College of Engineering

Mumbai, India mehtadhruv933@gmail.com

 

Jinay Parekh

Dept. of Computer Engineering Dwarkadas. J. Sanghvi College of Engineering

Mumbai, India jinayparekh2001@gmail.com

 

Pratik Kanani

Dept. of Computer Engineering

Dwarkadas. J. Sanghvi College of Engineering

Mumbai, India pratikkanani123@gmail.com

 

Harvy Gandhi

Dept. of Computer Engineering Dwarkadas. J. Sanghvi College of Engineering

Mumbai, India harvygandhi2@gmail.com

 

Tanya Sharma

Dept. of Electronics Engineering Dwarkadas. J. Sanghvi College of Engineering

Mumbai, India tanyasharma130599@gmail.com

 

Aniket Kore

Dept. of Computer Engineering

Dwarkadas. J. Sanghvi College of Engineering

Mumbai, India aniketkore24@gmail.com

Abstract—Content creators throughout the globe are con- fronted with an issue that cannot be solved by creativity. In today’s competitive marketing landscape, ensuring that your product is advertised to the correct audience in the proper direction is critical. A video Indexer combines multiple audio and video Artificial Intelligence (AI) technologies into a single integrated service, simplifying development. This paper presents a clean, one-stop solution that will instantly assist anyone wishing to sell their product by abstracting away all the dirty API calls and loading windows. This is done by offering three core services: label identification (video analysis), keyword identification (audio analysis), and sentiment analysis. Combining insights from these channels provides a list of Reddit communities that the user can then target to market their content effectively. Furthermore, by recognizing the sentiments inside the video, the user has an advantage by acquiring specific target markets for where to distribute their content. Additionally, intelligent emotion filtering allows users to customize the message’s tone.

Index Terms—LDA, SpaCy, Sentiment Analysis, NLTK Vader, Deep Face