http://repository.iiitd.edu.in/xmlui/handle/123456789/1810 Year-2025 Wed, 02 Sep 2026 08:35:32 GMT 2026-09-02T08:35:32Z http://repository.iiitd.edu.in/xmlui/handle/123456789/2072 Title: Designing Tactile graphics for the blind and visually impaired Authors: Verma, Ashwin; Gupta, Richa (Advisor) Abstract: This project builds upon the previous version of TacTile, which was developed in the prior semester using an EfficientNetV2-B0 and YOLO-based backbone. In the current version, we transitioned to using the MediaPipe Hand Landmark detector combined with a plain JSON file to store coordinates of the desired regions. To improve the robustness of the application, we have now trained a YOLOv8 model specifi- cally on cropped map regions. This enhancement ensures that the application functions reliably regardless of the phone’s height or position, as long as it remains parallel to the map surface. We are still using mediapipe for Hand Landmark ditector. Fri, 25 Jul 2025 00:00:00 GMT http://repository.iiitd.edu.in/xmlui/handle/123456789/2072 2025-07-25T00:00:00Z http://repository.iiitd.edu.in/xmlui/handle/123456789/2050 Title: Dynamic neural networks for continual learning Authors: Ghosh, Ratnango; Gupta, Vaibhav; Bagler, Ganesh (Advisor) Abstract: This report proposes a novel ensemble pipeline designed to address adaptive streaming natural language processing (NLP) challenges while mitigating catastrophic forgetting. Our architecture integrates five core components: dynamic slang detection, fruit fly–inspired sparse encoding, a memory-augmented streaming transformer, dual continual-learning defenses (Elastic Weight Consolidation and Unlabeled Knowledge Distillation), and an associative output layer. The pipeline operates across multiple temporal scales—short-term attention sinks and sliding windows, medium-term reservoir sampling, and long-term stable knowledge preservation—enabling efficient handling of real-time, evolving text streams. Key results from preliminary simulations indicate a 75% reduction in forgetting degradation, 20× speedup on long sequence processing, and 90% retention of prior-task performance after domain shifts. Our findings demonstrate the feasibility of layered defenses against forgetting and real-time adaptation to linguistic drift, offering robust performance for applications in chatbots, social media monitoring, and information retrieval. Fri, 18 Jul 2025 00:00:00 GMT http://repository.iiitd.edu.in/xmlui/handle/123456789/2050 2025-07-18T00:00:00Z