Shubhi Gupta

car-damage-assessment. Computer Vision and Deep Learning techniques to accurately classify vehicle damage to facilitate claims triage by training convolution neural networks

nlp-few-shot-relation-extraction. This project investigates few-shot learning for relation extraction using the FewRel dataset. We will compare Prototypical Networks, MAML, and k-NNs in different few-shot settings to see which performs best with minimal data. The goal is to improve relation extraction in NLP by effectively handling data scarcity.

satellite-image-stitching. Application of Big Data in Digital Image Processing. Using Apache Hadoop MapReduce with OpenCV for stitching of observatory tile images for object detection and recognition.

Direction-Detector-OpenCV. the following code is to detect the direction of arrows without using Hough Transformation. The code is in OpenCV Python and works only for LEFT, RIGHT, UP and DOWN directions.

image-recognition-with-AR-for-a-smart-in-store-shopping-experience. The aim of the project was to design a system to streamline the retail industry and enhance shopping experience. Using state of the art technology and AR Capabilities we were able to achieve a scope of automating the role of a salesperson. Customers can scan a product from a handheld device and obtain product information and visuals that assist the customer in the decision making process. The project was delivered under the guidance of Dr. B. Amrutha (Head Of Department CSE)

big-rag-theory. Python