May 2018 – Jul 2018 · 2 mos · Bengaluru Area, India

• Attention based methods for Attribute Detection :
• Studied various state-of-the-art methods for fine-grained classifcation focusing on diferent parts of the image.
• Implemented Multi-Attention CNN architecture with custom loss layers for attention maps in Tensorfow.
• The network can be used in a commercial solution for detecting fne-grained attributes in fashion/clothing.
• Got an accuracy of 76% on CUB200 bird dataset after experimenting with various networks like VGG19 & MobileNet.
• Awarded a Pre-Placement Ofer for the valuable contribution made in taking the project forward.

May 2017 – Jul 2017 · 2 mos · New Delhi Area, India

• Machine Learning Intern :
• Studied about Collaborative Filtering to give product recommendations to the users and implemented it in Python.
• Implemented Inventory Forecasting using anonymized data (of over 3 million users) on customer orders to predict
• which products user will buy again. Participated in the Instacart Market Basket Analysis competition held on Kaggle.
• Used the XGBoost package for training and got mean-F1 score of 0.386. Ranked among top 20% on the leaderboard.
• Software Developer Intern :
• Developed a mobile application using React Native for Android and iOS for connecting various sellers around Delhi.
• Integrated Google Firebase with react-native-gifted-chat package to make a chat console for easy communication
• between sellers and admin. The chat supports sending messages & images and stores them in the database