Niraj K. Jha
YOU?
Author Swipe
View article: SweetDeep: A Wearable AI Solution for Real-Time Non-Invasive Diabetes Screening
SweetDeep: A Wearable AI Solution for Real-Time Non-Invasive Diabetes Screening Open
The global rise in type 2 diabetes underscores the need for scalable and cost-effective screening methods. Current diagnosis requires biochemical assays, which are invasive and costly. Advances in consumer wearables have enabled early expl…
View article: First documented case of flunixin residue in a Himalayan Vulture <Gyps himalayensis> Hume, 1869 (Aves: Accipitriformes: Accipitridae) in India: conservation and veterinary implications
First documented case of flunixin residue in a Himalayan Vulture Hume, 1869 (Aves: Accipitriformes: Accipitridae) in India: conservation and veterinary implications Open
Non-steroidal anti-inflammatory drugs (NSAID), particularly diclofenac, have been widely identified as a major cause of vulture deaths across Asia, leading to significant population declines. The impact of other veterinary NSAIDs, includin…
View article: LinGen-Uni: A Universal Linear-Complexity Framework for High-Resolution Minute-Length Text-to-Video Generation
LinGen-Uni: A Universal Linear-Complexity Framework for High-Resolution Minute-Length Text-to-Video Generation Open
Text-to-video generation enhances content creation but is highly computationally intensive: The computational cost of Diffusion Transformers (DiTs) scales quadratically in the number of pixels. This makes minute-length video generation ext…
View article: Bottom-up Domain-specific Superintelligence: A Reliable Knowledge Graph is What We Need
Bottom-up Domain-specific Superintelligence: A Reliable Knowledge Graph is What We Need Open
Language models traditionally used for cross-domain generalization have recently demonstrated task-specific reasoning. However, their top-down training approach on general corpora is insufficient for acquiring abstractions needed for deep …
View article: TAD-SIE: sample size estimation for clinical randomized controlled trials using a Trend-Adaptive Design with a Synthetic-Intervention-Based Estimator
TAD-SIE: sample size estimation for clinical randomized controlled trials using a Trend-Adaptive Design with a Synthetic-Intervention-Based Estimator Open
View article: LLM Assisted Anomaly Detection Service for Site Reliability Engineers: Enhancing Cloud Infrastructure Resilience
LLM Assisted Anomaly Detection Service for Site Reliability Engineers: Enhancing Cloud Infrastructure Resilience Open
This paper introduces a scalable Anomaly Detection Service with a generalizable API tailored for industrial time-series data, designed to assist Site Reliability Engineers (SREs) in managing cloud infrastructure. The service enables effici…
View article: LinGen: Towards High-Resolution Minute-Length Text-to-Video Generation with Linear Computational Complexity
LinGen: Towards High-Resolution Minute-Length Text-to-Video Generation with Linear Computational Complexity Open
Text-to-video generation enhances content creation but is highly computationally intensive: The computational cost of Diffusion Transformers (DiTs) scales quadratically in the number of pixels. This makes minute-length video generation ext…
View article: Indian Leopard predation on the sub-adult Himalayan Griffon Vulture (Accipitridae: Accipitriformes)
Indian Leopard predation on the sub-adult Himalayan Griffon Vulture (Accipitridae: Accipitriformes) Open
This study documents instances of predation on Himalayan Griffon Vulture Gyps himalayansis by Indian Leopards Panthera pardus fusca Meyer, 1794 near the pre-release aviary of the Buxa Vulture Conservation Breeding Centre within the Buxa Ti…
View article: Process–Material–Performance Trade-off Exploration of Materials Sintering with Machine Learning Models
Process–Material–Performance Trade-off Exploration of Materials Sintering with Machine Learning Models Open
Process-induced porosity, defects, and residual stresses lead to mechanical performance degradation in fiber-reinforced composite and other heterogeneous structures. Physical and chemical processes create complex process–material–performan…
View article: COMFORT: A Continual Fine-Tuning Framework for Foundation Models Targeted at Consumer Healthcare
COMFORT: A Continual Fine-Tuning Framework for Foundation Models Targeted at Consumer Healthcare Open
Wearable medical sensors (WMSs) are revolutionizing smart healthcare by enabling continuous, real-time monitoring of user physiological signals, especially in the field of consumer healthcare. The integration of WMSs and modern machine lea…
View article: DOCTOR: A Multi-Disease Detection Continual Learning Framework Based on Wearable Medical Sensors
DOCTOR: A Multi-Disease Detection Continual Learning Framework Based on Wearable Medical Sensors Open
Modern advances in machine learning (ML) and wearable medical sensors (WMSs) in edge devices have enabled ML-driven disease detection for smart healthcare. Conventional ML-driven methods for disease detection rely on customizing individual…
View article: Learning Interpretable Differentiable Logic Networks
Learning Interpretable Differentiable Logic Networks Open
The ubiquity of neural networks (NNs) in real-world applications, from healthcare to natural language processing, underscores their immense utility in capturing complex relationships within high-dimensional data. However, NNs come with not…
View article: METRIK: Measurement-Efficient Randomized Controlled Trials using Transformers with Input Masking
METRIK: Measurement-Efficient Randomized Controlled Trials using Transformers with Input Masking Open
Clinical randomized controlled trials (RCTs) collect hundreds of measurements spanning various metric types (e.g., laboratory tests, cognitive/motor assessments, etc.) across 100s-1000s of subjects to evaluate the effect of a treatment, bu…
View article: CONFINE: Conformal Prediction for Interpretable Neural Networks
CONFINE: Conformal Prediction for Interpretable Neural Networks Open
Deep neural networks exhibit remarkable performance, yet their black-box nature limits their utility in fields like healthcare where interpretability is crucial. Existing explainability approaches often sacrifice accuracy and lack quantifi…
View article: Attention-Driven Training-Free Efficiency Enhancement of Diffusion Models
Attention-Driven Training-Free Efficiency Enhancement of Diffusion Models Open
Diffusion Models (DMs) have exhibited superior performance in generating high-quality and diverse images. However, this exceptional performance comes at the cost of expensive architectural design, particularly due to the attention module h…
View article: DynaMo: Accelerating Language Model Inference with Dynamic Multi-Token Sampling
DynaMo: Accelerating Language Model Inference with Dynamic Multi-Token Sampling Open
Traditional language models operate autoregressively, i.e., they predict one token at a time. Rapid explosion in model sizes has resulted in high inference times. In this work, we propose DynaMo, a suite of multi-token prediction language …
View article: PAGE: Domain-Incremental Adaptation with Past-Agnostic Generative Replay for Smart Healthcare
PAGE: Domain-Incremental Adaptation with Past-Agnostic Generative Replay for Smart Healthcare Open
We propose PAGE, a domain-incremental adaptation strategy with past-agnostic generative replay for smart healthcare. PAGE enables generative replay without the aid of any preserved data or information from prior domains. When adapting to a…
View article: SECRETS: Subject-efficient clinical randomized controlled trials using synthetic intervention
SECRETS: Subject-efficient clinical randomized controlled trials using synthetic intervention Open
View article: Neural Slot Interpreters: Grounding Object Semantics in Emergent Slot Representations
Neural Slot Interpreters: Grounding Object Semantics in Emergent Slot Representations Open
Several accounts of human cognition posit that our intelligence is rooted in our ability to form abstract composable concepts, ground them in our environment, and reason over these grounded entities. This trifecta of human thought has rema…
View article: TAD-SIE: Sample Size Estimation for Clinical Randomized Controlled Trials using a Trend-Adaptive Design with a Synthetic-Intervention-Based Estimator
TAD-SIE: Sample Size Estimation for Clinical Randomized Controlled Trials using a Trend-Adaptive Design with a Synthetic-Intervention-Based Estimator Open
Phase-3 clinical trials provide the highest level of evidence on drug safety and effectiveness needed for market approval by implementing large randomized controlled trials (RCTs). However, 30-40% of these trials fail mainly because such s…
View article: How MagNet: Machine Learning Framework for Modeling Power Magnetic Material Characteristics
How MagNet: Machine Learning Framework for Modeling Power Magnetic Material Characteristics Open
This paper applies machine learning to power magnetics modeling. We first introduce an open-source database – MagNet – which hosts a large amount of experimentally measured excitation data for many materials across a variety of operating c…
View article: SCouT: Synthetic Counterfactuals via Spatiotemporal Transformers for Actionable Healthcare
SCouT: Synthetic Counterfactuals via Spatiotemporal Transformers for Actionable Healthcare Open
The synthetic control method has pioneered a class of powerful data-driven techniques to estimate the counterfactual reality of a unit from donor units. At its core, the technique involves a linear model fitted on the pre-intervention peri…
View article: BREATHE: Second-Order Gradients and Heteroscedastic Emulation based Design Space Exploration
BREATHE: Second-Order Gradients and Heteroscedastic Emulation based Design Space Exploration Open
Researchers constantly strive to explore larger and more complex search spaces in various scientific studies and physical experiments. However, such investigations often involve sophisticated simulators or time-consuming experiments that m…
View article: Why MagNet: Quantifying the Complexity of Modeling Power Magnetic Material Characteristics
Why MagNet: Quantifying the Complexity of Modeling Power Magnetic Material Characteristics Open
This paper motivates the development of sophisticated data–driven models for power magnetic material characteristics.Core losses and hysteresis loops are critical information in the design process of power magnetics, yet the physics behind…
View article: Zero-TPrune: Zero-Shot Token Pruning through Leveraging of the Attention Graph in Pre-Trained Transformers
Zero-TPrune: Zero-Shot Token Pruning through Leveraging of the Attention Graph in Pre-Trained Transformers Open
Deployment of Transformer models on edge devices is becoming increasingly challenging due to the exponentially growing inference cost that scales quadratically with the number of tokens in the input sequence. Token pruning is an emerging s…
View article: Im-Promptu: In-Context Composition from Image Prompts
Im-Promptu: In-Context Composition from Image Prompts Open
Large language models are few-shot learners that can solve diverse tasks from a handful of demonstrations. This implicit understanding of tasks suggests that the attention mechanisms over word tokens may play a role in analogical reasoning…
View article: DOCTOR: A Multi-Disease Detection Continual Learning Framework Based on Wearable Medical Sensors
DOCTOR: A Multi-Disease Detection Continual Learning Framework Based on Wearable Medical Sensors Open
Modern advances in machine learning (ML) and wearable medical sensors (WMSs) in edge devices have enabled ML-driven disease detection for smart healthcare. Conventional ML-driven methods for disease detection rely on customizing individual…
View article: SECRETS: Subject-Efficient Clinical Randomized Controlled Trials using Synthetic Intervention
SECRETS: Subject-Efficient Clinical Randomized Controlled Trials using Synthetic Intervention Open
The randomized controlled trial (RCT) is the gold standard for estimating the average treatment effect (ATE) of a medical intervention but requires 100s-1000s of subjects, making it expensive and difficult to implement. While a cross-over …
View article: FlexiBERT: Are Current Transformer Architectures too Homogeneous and Rigid?
FlexiBERT: Are Current Transformer Architectures too Homogeneous and Rigid? Open
The existence of a plethora of language models makes the problem of selecting the best one for a custom task challenging. Most state-of-the-art methods leverage transformer-based models (e.g., BERT) or their variants. However, training suc…
View article: TransCODE: Co-design of Transformers and Accelerators for Efficient Training and Inference
TransCODE: Co-design of Transformers and Accelerators for Efficient Training and Inference Open
Automated co-design of machine learning models and evaluation hardware is critical for efficiently deploying such models at scale. Despite the state-of-the-art performance of transformer models, they are not yet ready for execution on reso…