arXiv subjects

John A. Stankovic

Publications and source records attributed to John A. Stankovic.

11 recordsLinked to original sources

arXiv2024

DKEC: Domain Knowledge Enhanced Multi-Label Classification for Diagnosis Prediction

Multi-label text classification (MLTC) tasks in the medical domain often face the long-tail label distribution problem. Prior works have explored hierarchical label structures to find relevant information for few-shot classes, but mostly neglected to incorporate external knowledge from medical guidelines. This paper presents DKEC, Domain Knowledge Enhanced Classification for diagnosis prediction with two innovations: (1) automated construction of heterogeneous knowledge graphs from external sources to capture semantic relations among diverse medical entities, (2) incorporating the heterogeneous knowledge graphs in few-shot classification using a label-wise attention mechanism. We construct DKEC using three online medical knowledge sources and evaluate it on a real-world Emergency Medical Services (EMS) dataset and a public electronic health record (EHR) dataset. Results show that DKEC outperforms the state-of-the-art label-wise attention networks and transformer models of different sizes, particularly for the few-shot classes. More importantly, it helps the smaller language models achieve comparable performance to large language models.

cs.CL↗

arXiv2024

Real-Time Multimodal Cognitive Assistant for Emergency Medical Services

Emergency Medical Services (EMS) responders often operate under time-sensitive conditions, facing cognitive overload and inherent risks, requiring essential skills in critical thinking and rapid decision-making. This paper presents CognitiveEMS, an end-to-end wearable cognitive assistant system that can act as a collaborative virtual partner engaging in the real-time acquisition and analysis of multimodal data from an emergency scene and interacting with EMS responders through Augmented Reality (AR) smart glasses. CognitiveEMS processes the continuous streams of data in real-time and leverages edge computing to provide assistance in EMS protocol selection and intervention recognition. We address key technical challenges in real-time cognitive assistance by introducing three novel components: (i) a Speech Recognition model that is fine-tuned for real-world medical emergency conversations using simulated EMS audio recordings, augmented with synthetic data generated by large language models (LLMs); (ii) an EMS Protocol Prediction model that combines state-of-the-art (SOTA) tiny language models with EMS domain knowledge using graph-based attention mechanisms; (iii) an EMS Action Recognition module which leverages multimodal audio and video data and protocol predictions to infer the intervention/treatment actions taken by the responders at the incident scene. Our results show that for speech recognition we achieve superior performance compared to SOTA (WER of 0.290 vs. 0.618) on conversational data. Our protocol prediction component also significantly outperforms SOTA (top-3 accuracy of 0.800 vs. 0.200) and the action recognition achieves an accuracy of 0.727, while maintaining an end-to-end latency of 3.78s for protocol prediction on the edge and 0.31s on the server.

cs.AI↗

arXiv2023

Camera-Independent Single Image Depth Estimation from Defocus Blur

Monocular depth estimation is an important step in many downstream tasks in machine vision. We address the topic of estimating monocular depth from defocus blur which can yield more accurate results than the semantic based depth estimation methods. The existing monocular depth from defocus techniques are sensitive to the particular camera that the images are taken from. We show how several camera-related parameters affect the defocus blur using optical physics equations and how they make the defocus blur depend on these parameters. The simple correction procedure we propose can alleviate this problem which does not require any retraining of the original model. We created a synthetic dataset which can be used to test the camera independent performance of depth from defocus blur models. We evaluate our model on both synthetic and real datasets (DDFF12 and NYU depth V2) obtained with different cameras and show that our methods are significantly more robust to the changes of cameras. Code: https://github.com/sleekEagle/defocus_camind.git

cs.CV↗

arXiv2023

CitySpec with Shield: A Secure Intelligent Assistant for Requirement Formalization

An increasing number of monitoring systems have been developed in smart cities to ensure that the real-time operations of a city satisfy safety and performance requirements. However, many existing city requirements are written in English with missing, inaccurate, or ambiguous information. There is a high demand for assisting city policymakers in converting human-specified requirements to machine-understandable formal specifications for monitoring systems. To tackle this limitation, we build CitySpec, the first intelligent assistant system for requirement specification in smart cities. To create CitySpec, we first collect over 1,500 real-world city requirements across different domains (e.g., transportation and energy) from over 100 cities and extract city-specific knowledge to generate a dataset of city vocabulary with 3,061 words. We also build a translation model and enhance it through requirement synthesis and develop a novel online learning framework with shielded validation. The evaluation results on real-world city requirements show that CitySpec increases the sentence-level accuracy of requirement specification from 59.02% to 86.64%, and has strong adaptability to a new city and a new domain (e.g., the F1 score for requirements in Seattle increases from 77.6% to 93.75% with online learning). After the enhancement from the shield function, CitySpec is now immune to most known textual adversarial inputs (e.g., the attack success rate of DeepWordBug after the shield function is reduced to 0% from 82.73%). We test the CitySpec with 18 participants from different domains. CitySpec shows its strong usability and adaptability to different domains, and also its robustness to malicious inputs.

cs.AI↗

arXiv2022

CitySpec: An Intelligent Assistant System for Requirement Specification in Smart Cities

An increasing number of monitoring systems have been developed in smart cities to ensure that real-time operations of a city satisfy safety and performance requirements. However, many existing city requirements are written in English with missing, inaccurate, or ambiguous information. There is a high demand for assisting city policy makers in converting human-specified requirements to machine-understandable formal specifications for monitoring systems. To tackle this limitation, we build CitySpec, the first intelligent assistant system for requirement specification in smart cities. To create CitySpec, we first collect over 1,500 real-world city requirements across different domains from over 100 cities and extract city-specific knowledge to generate a dataset of city vocabulary with 3,061 words. We also build a translation model and enhance it through requirement synthesis and develop a novel online learning framework with validation under uncertainty. The evaluation results on real-world city requirements show that CitySpec increases the sentence-level accuracy of requirement specification from 59.02% to 86.64%, and has strong adaptability to a new city and a new domain (e.g., F1 score for requirements in Seattle increases from 77.6% to 93.75% with online learning).

cs.AI↗

arXiv2021

iWash: A Smartwatch Handwashing Quality Assessment and Reminder System with Real-time Feedback in the Context of Infectious Disease

Washing hands properly and frequently is the simplest and most cost-effective interventions to prevent the spread of infectious diseases. People are often ignorant about proper handwashing in different situations and do not know if they wash hands properly. Smartwatches are found to be effective for assessing the quality of handwashing. However, the existing smartwatch based systems are not comprehensive enough in terms of achieving accuracy as well as reminding people to handwash and providing feedback to the user about the quality of handwashing. On-device processing is often required to provide real-time feedback to the user, and so it is important to develop a system that runs efficiently on low-resource devices like smartwatches. However, none of the existing systems for handwashing quality assessment are optimized for on-device processing. We present iWash, a comprehensive system for quality assessment and context-aware reminder for handwashing with real-time feedback using smartwatches. iWash is a hybrid deep neural network based system that is optimized for on-device processing to ensure high accuracy with minimal processing time and battery usage. Additionally, it is a context-aware system that detects when the user is entering home using a Bluetooth beacon and provides reminders to wash hands. iWash also offers touch-free interaction between the user and the smartwatch that minimizes the risk of germ transmission. We collected a real-life dataset and conducted extensive evaluations to demonstrate the performance of iWash. Compared to the existing handwashing quality assessment systems, we achieve around 12% higher accuracy for quality assessment, as well as we reduce the processing time and battery usage by around 37% and 10%, respectively.

cs.HC↗

arXiv2020

MFED: A System for Monitoring Family Eating Dynamics

Obesity is a risk factor for many health issues, including heart disease, diabetes, osteoarthritis, and certain cancers. One of the primary behavioral causes, dietary intake, has proven particularly challenging to measure and track. Current behavioral science suggests that family eating dynamics (FED) have high potential to impact child and parent dietary intake, and ultimately the risk of obesity. Monitoring FED requires information about when and where eating events are occurring, the presence or absence of family members during eating events, and some person-level states such as stress, mood, and hunger. To date, there exists no system for real-time monitoring of FED. This paper presents MFED, the first of its kind of system for monitoring FED in the wild in real-time. Smart wearables and Bluetooth beacons are used to monitor and detect eating activities and the location of the users at home. A smartphone is used for the Ecological Momentary Assessment (EMA) of a number of behaviors, states, and situations. While the system itself is novel, we also present a novel and efficient algorithm for detecting eating events from wrist-worn accelerometer data. The algorithm improves eating gesture detection F1-score by 19% with less than 20% computation compared to the state-of-the-art methods. To date, the MFED system has been deployed in 20 homes with a total of 74 participants, and responses from 4750 EMA surveys have been collected. This paper describes the system components, reports on the eating detection results from the deployments, proposes two techniques for improving ground truth collection after the system is deployed, and provides an overview of the FED data, generated from the multi-component system, that can be used to model and more comprehensively understand insights into the monitoring of family eating dynamics.

cs.CY↗

arXiv2020

See Through Smoke: Robust Indoor Mapping with Low-cost mmWave Radar

This paper presents the design, implementation and evaluation of milliMap, a single-chip millimetre wave (mmWave) radar based indoor mapping system targetted towards low-visibility environments to assist in emergency response. A unique feature of milliMap is that it only leverages a low-cost, off-the-shelf mmWave radar, but can reconstruct a dense grid map with accuracy comparable to lidar, as well as providing semantic annotations of objects on the map. milliMap makes two key technical contributions. First, it autonomously overcomes the sparsity and multi-path noise of mmWave signals by combining cross-modal supervision from a co-located lidar during training and the strong geometric priors of indoor spaces. Second, it takes the spectral response of mmWave reflections as features to robustly identify different types of objects e.g. doors, walls etc. Extensive experiments in different indoor environments show that milliMap can achieve a map reconstruction error less than 0.2m and classify key semantics with an accuracy around 90%, whilst operating through dense smoke.

eess.SP↗

arXiv2019

Hardware/Software Security Patches for Internet of Trillions of Things

With the rapid development of the Internet of Things, there are many interacting devices and applications. One crucial challenge is how to provide security. Our proposal for a new direction is to create "smart buttons" and collections of them called "smart blankets" as hardware/software security patches rather than software-only patches.

cs.CR↗

arXiv2016

Systems Computing Challenges in the Internet of Things

A recent McKinsey report estimates the economic impact of the Internet of Things (IoT) to be between $3.9 to $11 trillion dollars by 20251 . IoT has the potential to have a profound impact on our daily lives, including technologies for the home, for health, for transportation, and for managing our natural resources. The Internet was largely driven by information and ideas generated by people, but advances in sensing and hardware have enabled computers to more easily observe the physical world. Coupling this additional layer of information with advances in machine learning brings dramatic new capabilities including the ability to capture and process tremendous amounts of data; to predict behaviors, activities, and the future in uncanny ways; and to manipulate the physical world in response. This trend will fundamentally change how people interact with physical objects and the environment. Success in developing value-added capabilities around IoT requires a broad approach that includes expertise in sensing and hardware, machine learning, networked systems, human-computer interaction, security, and privacy. Strategies for making IoT practical and spurring its ultimate adoption also require a multifaceted approach that often transcends technology, such as with concerns over data security, privacy, public policy, and regulatory issues. In this paper we argue that existing best practices in building robust and secure systems are insufficient to address the new challenges that IoT systems will present. We provide recommendations regarding investments in research areas that will help address inadequacies in existing systems, practices, tools, and policies.

cs.CY↗

arXiv2016

Taxi Dispatch with Real-Time Sensing Data in Metropolitan Areas: A Receding Horizon Control Approach

Traditional taxi systems in metropolitan areas often suffer from inefficiencies due to uncoordinated actions as system capacity and customer demand change. With the pervasive deployment of networked sensors in modern vehicles, large amounts of information regarding customer demand and system status can be collected in real time. This information provides opportunities to perform various types of control and coordination for large-scale intelligent transportation systems. In this paper, we present a receding horizon control (RHC) framework to dispatch taxis, which incorporates highly spatiotemporally correlated demand/supply models and real-time GPS location and occupancy information. The objectives include matching spatiotemporal ratio between demand and supply for service quality with minimum current and anticipated future taxi idle driving distance. Extensive trace-driven analysis with a data set containing taxi operational records in San Francisco shows that our solution reduces the average total idle distance by 52%, and reduces the supply demand ratio error across the city during one experimental time slot by 45%. Moreover, our RHC framework is compatible with a wide variety of predictive models and optimization problem formulations. This compatibility property allows us to solve robust optimization problems with corresponding demand uncertainty models that provide disruptive event information.

eess.SY↗