Deep Learning Approach for Hand Drawn Emoji Identification

Object detection is a computer technology which is based on image processing and deep learning. It uses a database to recognize objects by comparing them. Emoticons are small symbols or pictures used to express opinions or emotions in instant messages. They are widely used on various social media platforms like Facebook, Twitter, Instagram, etc. In this paper, we focus on hand-drawn emoticons and classify them into 8 categories. Hand-drawn emoticons are created using advanced technology or drawn by hand with a pen. This paper aims to help users organize these emoticons for use in various social media with minimal confusion. We created a dataset of 4000 hand-drawn emoticon images, with 500 images per category, to train our model. Our framework uses a convolutional neural network model to recognize and classify the emoticons with a precision of 64%.