ANRF–PAIR Project
Progress Reports
Status updates on active research tasks, covering objectives, methodology, and outcomes to date.
| Project Title | PI / Co-PI | Objective | Methodology | Progress |
|---|---|---|---|---|
| Regional Language Sentiment Analysis Using Transformer Models |
PI Dr. Meena Iyer
Co-PI Ishaan Gupta
|
Build and evaluate a fine-tuned multilingual transformer model for code-mixed sentiment classification. | Collected and annotated 50,000 social media posts across three regional languages and fine-tuned a mBERT-based classifier. | Achieved 89% F1-score on the held-out validation set. Preparing manuscript for submission to a peer-reviewed NLP venue. |
| Biocompatible Hydrogels for Targeted Drug Delivery |
PI Dr. Aris Thorne
|
Synthesize and characterize pH-responsive hydrogel matrices for tumor-targeted drug release. | Utilized cross-linking polymerization methods and characterized swelling ratios using thermal gravimetric analysis. | Synthesis phase complete. Initial degradation trials indicate steady release kinetics under acidic conditions matching targeted tumor microenvironments. |
| Autonomous Drone Navigation via Deep Reinforcement Learning |
PI Prof. Elena Rostova
Co-PI Dr. Marcus Vance
|
Train collision avoidance models in high-density simulated urban environments. | Deployed Proximal Policy Optimization (PPO) agents within a simulated Unity environment using synthetic LiDAR inputs. | Agents have achieved a 94% success rate in avoiding dynamic obstacles. Currently working on physical deployment validation protocols. |