Research Projects
Evolving Mixture of Low-Rank Experts for Continual User Modeling Zillow Group
- Developed a mixture-of-rank-1-experts architecture for continual user modeling, enabling parameter-efficient adaptation to sequential recommendation data.
- Paper published in ECIR 2026.
Bayesian Network Structure Adaptation for Continual Learning RIT
- Developed a continual structure adaptation framework that integrates beta-Bernoulli processes for structure inference within the sequential Bayes framework, enabling dynamic evolution of both network depth and width in continual learning scenarios.
- Paper published in ICML 2024.
Cross-Task Representation Alignment for Exemplar-Free Class-Incremental Learning RIT
- Developed a cross-task representation-alignment framework that improved average accuracy by 2.16 percentage points for exemplar-free class-incremental learning.
- Manuscript under review.
Bayesian Adaptive Graph Neural Network for Gene-Disease Association RIT
- Developing a Bayesian adaptive graph neural network for gene-disease association prediction over protein-protein interaction graphs.
- Currently conducting benchmark evaluation.
Nepalese License Plate Recognition, Undergraduate Capstone Project Pulchowk Campus
- Developed a license plate recognition system, tailored for Nepalese license plates, with three key stages: vehicle detection, license plate localization, and Nepalese character recognition.
- Created a license plate localization dataset by annotating Nepalese vehicle images, alongside building a Nepalese character recognition dataset using Devanagari fonts.
Industry Projects
Human Trafficking Recognition from Online Advertisements and Inter-Ad Matching Fusemachines
- Led the development of a multimodal machine-learning pipeline for identifying potential trafficking activity in online advertisements using video, image, and text data.
- Built image–text contrastive models for advertisement matching, face-based identity linking, and BERT-based social-handle extraction, improving cross-ad linkage accuracy by 35%.
Waste Type Detection Fusemachines
- Developed a lightweight object detector that increased inference throughput by 47%, then deployed it on NVIDIA Jetson Nano devices for real-time waste-type and disposal-intent classification.
Analysis of Radio Panelists Data Fusemachines
- Analyzed the impact of song quality, commercial length, and time of day on panelist retention, designing custom metrics and statistical tests to quantify song quality.
- Resolved date-inconsistency bugs in the existing data pipeline and built a feature-engineering pipeline to augment the dataset.
Session-based Network Intrusion Detection System Fusemachines
- Designed and validated an AutoEncoder-based semi-supervised learning pipeline for network anomaly detection, engineering session-level features from raw pcap traffic captures.
