9 papers
Classifier-Dependent Benefits of Pseudo-Labeling for Semi-Supervised Android Malware Attribution
Md Rafid Islam, Zahid Hasan, Hafiz Abdur Rahman
Detecting and classifying Android malware families remains challenging due to high feature dimensionality, class imbalance, and the high cost of expert-labeled data. Semi-supervise…
Exploration-Driven Personalized Federated Reinforcement Learning via Intrinsic Motivation
Md Rafid Islam, Rafsan Jany, Zahid Hasan +1
Personalized Federated Reinforcement Learning (PFRL) takes a decentralized approach to storing and accessing information based on past experiences while keeping each client's data…
Contactless Respiratory Monitoring on Heterogeneous Mobile Robots: A Multimodal Edge-Computing Framework
Milind Rampure, Shadman Sakib, Haley Patel +2
Respiratory-rate (RR) monitoring is a critical component of remote triage and victim assessment in emergency response, disaster recovery, and infectious-disease scenarios, where mi…
Lorentz Framework for Semantic Segmentation
Zahid Hasan, Masud Ahmed, Nirmalya Roy
Semantic segmentation in hyperbolic space enables compact modeling of hierarchical structure while providing inherent uncertainty quantification. Prior approaches predominantly rel…
CAM-Seg: A Continuous-valued Embedding Approach for Semantic Image Generation
Masud Ahmed, Zahid Hasan, Syed Arefinul Haque +4
Traditional transformer-based semantic segmentation relies on quantized embeddings. However, our analysis reveals that autoencoder accuracy on segmentation mask using quantized emb…
Novel Categories Discovery Via Constraints on Empirical Prediction Statistics
Zahid Hasan, Abu Zaher Md Faridee, Masud Ahmed +4
Novel Categories Discovery (NCD) aims to cluster novel data based on the class semantics of known classes using the open-world partial class space annotated dataset. As an alternat…
A Systematic Study on Object Recognition Using Millimeter-wave Radar
Maloy Kumar Devnath, Avijoy Chakma, Mohammad Saeid Anwar +5
Due to its light and weather-independent sensing, millimeter-wave (MMW) radar is essential in smart environments. Intelligent vehicle systems and industry-grade MMW radars have int…
NEV-NCD: Negative Learning, Entropy, and Variance regularization based novel action categories discovery
Zahid Hasan, Masud Ahmed, Abu Zaher Md Faridee +4
Novel Categories Discovery (NCD) facilitates learning from a partially annotated label space and enables deep learning (DL) models to operate in an open-world setting by identifyin…
Demo: RhythmEdge: Enabling Contactless Heart Rate Estimation on the Edge
Zahid Hasan, Emon Dey, Sreenivasan Ramasamy Ramamurthy +2
In this demo paper, we design and prototype RhythmEdge, a low-cost, deep-learning-based contact-less system for regular HR monitoring applications. RhythmEdge benefits over existin…