A5008699846- Image and Video Quality Assessment
- Data Stream Mining Techniques
- Video Coding and Compression Technologies
- Semantic Web and Ontologies
- Advanced Chemical Sensor Technologies
- Intravenous Infusion Technology and Safety
- Hand Gesture Recognition Systems
- Air Quality Monitoring and Forecasting
- Additive Manufacturing and 3D Printing Technologies
- Machine Fault Diagnosis Techniques
- Industrial Vision Systems and Defect Detection
- Smart Grid Energy Management
- Water Quality Monitoring Technologies
- Caching and Content Delivery
- IoT and Edge/Fog Computing
- Manufacturing Process and Optimization
- Visual Attention and Saliency Detection
- Underwater Vehicles and Communication Systems
- Image and Object Detection Techniques
- Machine Learning and Data Classification
- Human Motion and Animation
- Multimedia Communication and Technology
- Optical Wireless Communication Technologies
- Virtual Reality Applications and Impacts
- Advanced Image Processing Techniques
University of Illinois Urbana-Champaign
2021-2024
Underwater communication is essential for both recreational and scientific activities, such as scuba diving. However, existing methods remain highly constrained by environmental challenges often require specialized hardware, driving research into more accessible underwater solutions. While recent acoustic-based systems support text messaging on mobile devices, their low data rates severely limit broader applications. We present AquaScope, the first acoustic system capable of image...
In the rapidly evolving landscape of scientific semiconductor laboratories (commonly known as, cleanrooms), integrated with Internet Things (IoT) technology and Cyber-Physical Systems (CPSs), several factors including operational changes, sensor aging, software updates introduction new processes or equipment can lead to dynamic non-stationary data distributions in streams. This phenomenon, as concept drift, poses a substantial challenge for traditional data-driven digital twin static machine...
Deep learning video analytic systems process live feeds from multiple cameras with computer vision models deployed on edge or cloud. To optimize utility for these systems, which usually corresponds to query accuracy, efficient bandwidth management the competing fluctuating network resources is crucial. We propose DeepStream, a multi-camera streaming system deep analytics. DeepStream addresses challenge of limited and by offering several tailored solutions. design novel Regions Interest...
360-degree video is becoming an integral part of our content consumption through both on demand and live broadcast services. However, still challenging due to the huge network bandwidth cost if all views are delivered a large viewer population over diverse networks. In this paper, we present 360BroadView, management approach viewport prediction in broadcast. We make some high-bandwidth viewers be leading help others (lagging viewers) predict viewports during viewing save bandwidth. Our...
Smaller-scale Vision-Langauge Models (VLMs) often claim to perform on par with larger models in general-domain visual grounding and question-answering benchmarks while offering advantages computational efficiency storage. However, their ability handle rare objects, which fall into the long tail of data distributions, is less understood. To rigorously evaluate this aspect, we introduce "Uncontextualized Uncommon Objects" (UOUO) benchmark. This benchmark focuses systematically testing VLMs...
Fused Deposition Modeling (FDM) is a widely used additive manufacturing (AM) technique valued for its flexibility and cost-efficiency, with applications in variety of industries including healthcare aerospace. Recent developments have made affordable FDM machines accessible encouraged adoption among diverse users. However, the design, planning, production process require specialized interdisciplinary knowledge. Managing complex parameters resolving print defects remain challenging. These...
Sensory IoT (Internet of Things) networks are widely applied and studied in recent years have demonstrated their unique benefits various areas. In this paper, we bring the sensor network to an application scenario that has rarely been - academic cleanrooms. We design SENSELET++, a low-cost sensing platform can collect, manage analyze large amount sensory data from heterogeneous sensors. Furthermore, novel hybrid anomaly detection framework which detect both time-critical complex non-critical...
Ultrasonic welding machines play a critical role in the lithium battery industry, facilitating bonding of batteries with conductors. Ensuring high-quality is vital, making tool condition monitoring systems essential for early-stage quality control. However, existing methods face challenges cost, downtime, and adaptability. In this paper, we present WeldMon, an affordable ultrasonic machine system that utilizes custom data acquisition analysis pipeline designed real-time analysis. Our...
Ultrasonic welding machines play a critical role in the lithium battery industry, facilitating bonding of batteries with conductors. Ensuring high-quality is vital, making tool condition monitoring systems essential for early-stage quality control. However, existing methods face challenges cost, downtime, and adaptability. In this paper, we present WeldMon, an affordable ultrasonic machine system that utilizes custom data acquisition analysis pipeline designed real-time analysis. Our...
The three-tier sense-edge-cloud (SEC) framework developed as a consequence of the advances in IoT communication represents an inverted multicast paradigm, allowing sensors to deliver data cloud via edge devices. Due increasing volume generation and transmission, SEC is often challenged by resource bottlenecks. To remedy this problem, network must exhibit three key characteristics: optimal efficiency for energy sustain applications, focus on fitness score through selective streaming. Taking...
The emergence of 360-video streaming systems has brought about new possibilities for immersive video experiences while requiring significantly higher bandwidth than traditional 2D streaming. Viewport prediction is used to address this problem, but interesting storylines outside the viewport are ignored. To limitation, we present SAVG360, a novel guidance system that utilizes global content information available on server side enhance with best saliency-captured storyline 360-videos. saliency...