Engineered sequence-based models, automated software testing, and developed computer vision algorithms to enhance traffic safety and detection of rule violators.
Drove innovation through scalable AI algorithms, automated deployment framework, and novel integration of physics-based simulation, resulting in significant cost savings and process optimization for clients.
Led user-centric product development, employing interviews, usability tests, and UI enhancements to deliver a highly-adaptable MVP, resulting in securing multimillion-euro contracts.
Headed the technical efforts of a team of 25 students in developing India's first fully autonomous car, securing significant funding and implementing state-of-the-art computer vision algorithms for real-time object detection and semantic segmentation.
Unleashed real-time Road-Lane semantic segmentation for autonomous vehicles with Linknet, crafting a robust codebase and seamless ROS integration.
Led a 10-member team to a Bronze Medal, creating an intuitive GUI for deep learning model training and visualization. Employed SMOTE, t-SNE, and Grad-CAM & LIME for enhanced results.
Embarked on an artistic journey with Text to Photo-realistic Image Synthesis using StackGAN. Crafted lifelike visuals from text descriptions and fine-tuned performance with Conditioning Augmentation.
Dived into the world of Object Detection in Thermal Images, tailoring Keras-Retinanet for precise detection of pedestrians and vehicles. Pioneered the fusion of LSTM and Retinanet, elevating the model's performance with sequential insights.
Thrived in the Spotify Sequential Skip Prediction Challenge, an exhilarating AIcrowd competition organized by Spotify. Crafted a cutting-edge model leveraging BiLSTM-based feature-encoder and encoder-decoder architecture to predict song skips using historical data.
Took charge in the ASME Student Design Challenge, leading a 22-member team to secure a global 4th place. Engineered an innovative, solar-powered carrier bot with navigation capabilities in the arena.