Final-year B.S. (Honors), Data Science & AI @ IIT Guwahati — working on sleep apnea classification, wearable fall detection, and computational phonetics.
|
ConCAD (Contrastive Cross-Attention Dual-branch) architecture fusing ECG, RRI, RPE & Mel spectrogram signals across PhysioNet Apnea-ECG, UCDDB, and MIT-BIH slpdb.
|
Cross-dataset CNN-LSTM comparison (SisFall vs FallAllD), LOSO CV. SisFall AUC: 98.81% · FallAllD AUC: 86.67%
|
|
Replicating and extending the humidity–vowel hypothesis (Everett 2017) with PHOIBLE 2.0, focused on Indian languages.
|
Applied machine learning project.
|
💬 "Learning never exhausts the mind." – Leonardo da Vinci
