A5071026265- CCD and CMOS Imaging Sensors
- Visual Attention and Saliency Detection
- Advanced Image and Video Retrieval Techniques
- Advanced Neural Network Applications
- Advanced Vision and Imaging
- Tactile and Sensory Interactions
- Industrial Vision Systems and Defect Detection
- Ultrasonics and Acoustic Wave Propagation
- Radio Frequency Integrated Circuit Design
- Optical measurement and interference techniques
- Image and Video Quality Assessment
- Millimeter-Wave Propagation and Modeling
- Non-Destructive Testing Techniques
- Advanced Memory and Neural Computing
- Microwave Engineering and Waveguides
- Structural Health Monitoring Techniques
- Smart Parking Systems Research
- Geophysical Methods and Applications
- Image Processing Techniques and Applications
- Context-Aware Activity Recognition Systems
- RFID technology advancements
- Video Surveillance and Tracking Methods
- Gaze Tracking and Assistive Technology
- Advanced Optical Sensing Technologies
- Video Coding and Compression Technologies
Samsung (United States)
2021-2024
Pennsylvania State University
2010-2016
Guidedwave (United States)
2011
It is projected that mobile traffic will increase by 80x year 2030. To meet this in demand, it inevitable to utilize the terahertz bands (0.1 THz 10 THz) for future 6G wireless systems. However, operating at such high frequency comes with several fundamental and technical challenges. In work, we present a proof-of-concept system demonstrate feasibility of establishing communication link 140 GHz carrier frequency. addition, work highlights techniques tackle challenges regime. authors...
According to the World Health Organization, "285 million people are estimated be visually impaired worldwide" [1]. Several technologies such as automatic text readers, Braille note makers, and navigation assistance canes have been developed assist impaired. Concurrent advances in computer vision hardware provide opportunities for a visual-assistance system that can used multiple contexts. As part of Visual Cortex on Silicon program, we developing interfaces, algorithms, platforms with focus...
6G is envisioned to provide ultimate experience for all through hyper-connectivity involving humans and everything, with unprecedented requirements expectations [1]. In this vision, terahertz (THz) technology a leading candidate realize the requirements. This paper presents latest development results of wireless prototyping platform, which being developed in Samsung research lab. The platform currently supports real-time transmission 6 Gbps data over 2 GHz channel centered around 135...
Limited power budgets and the need for high performance computing have led to platform customization with a number of accelerators integrated CMPs. In order study customized architectures, we model four design points compare their energy across computer vision workloads. We analyze limitations generic architectures quantify costs increasing using these micro-architectural points. This analysis leads us develop framework consisting low-power multi-cores an array configurable micro-accelerator...
In any visual object recognition system, the classification accuracy will likely determine usefulness of system as a whole. many real-world applications, it is also important to be able recognize large number diverse objects for robust enough handle sort tasks that human handles on an average day. These objectives are often at odds with performance, running too detectors one scene prohibitively slow use in real-time scenario. However, information has temporal and spatial context can...
The Human Visual System (HVS) exhibits multi-resolution characteristics, where the fovea is at highest resolution while tapers off towards periphery. Given enough activity periphery, HVS then capable to foveate next region of interest (ROI), attend it full resolution. Saliency models in past have focused on identifying features that can be used a bottom-up manner generate conspicuity maps, which are combined together provide regions fixated interest. However, these neglect take into...
As high-fidelity small form-factor cameras become increasingly available and affordable, there will be a subsequent growth emergence of vision-based applications that take advantage this increase in visual information. The key challenge is for the embedded systems, on which bulk these deployed, to maintain real-time performance midst exponential spatial temporal data. For example, useful driver assistance system needs locate identify critical objects such as pedestrians, other vehicles,...
Video applications are becoming ubiquitous in mobile and embedded systems. Wearable video systems such as Google Glasses require capabilities for real-time analytics prolonged battery lifetimes. Further, the increasing resolution of image sensors these places an demand on both memory storage well computational power. In this work, we present Refresh Enabled Analytics (REVA) system, architecture multi-object scene understanding tackle unique opportunities provided by reducing DRAM refresh...
Embedded visual assist systems are emerging as increasingly viable tools for aiding visually impaired persons in their day-to-day life activities. Novel wearable devices with imaging capabilities will be uniquely positioned to activities such grocery shopping. However, supporting time-sensitive applications on embedded platforms requires an intelligent trade-off between accuracy and computational efficiency. In order maximize utility real-world scenarios, classifiers often need recognize...
Predicting dense depth accurately is essential for 3D scene understanding applications such as autonomous driving and robotics. However, the obtained from commercially available LiDAR Time-of-Flight sensors very sparse. With RGB color guidance, modern convolutional neural network (CNN) based approaches can recover missing information. there could be scenarios low-light environments where it might difficult to get an associated image with sparse depth. In this work, we propose a Generative...
Embedded visual assist systems are emerging as increasingly viable tools for aiding visually impaired persons in their day-to-day life activities. Novel wearable devices with imaging capabilities will be uniquely positioned to activities such grocery shopping. However, supporting time-sensitive applications on embedded platforms requires an intelligent trade-off between accuracy and computational efficiency. In order maximize utility real-world scenarios, classifiers often need recognize...
Ultrasonic guided waves are now proving to be a viable method for real-world long-range Non-Destructive Evaluation (NDE) applications. In order generate specific wave mode optimally, knowledge of the sensor parameters becomes imperative. This paper attempts experimentally measure beam divergence in an Electro-Magnetic Acoustic Transducer (EMAT) that is used Shear Horizontal (SH) mild steel plate. The commercial finite element package ABAQUS™ then run 3D simulations validate these...
Accurately measuring the absolute depth of every pixel captured by an imaging sensor is critical importance in real-time applications such as autonomous navigation, augmented reality and robotics. In order to predict dense depth, a general approach fuse inputs from different modalities LiDAR, camera other time-of-flight sensors. LiDAR sensors provide accurate data but are quite sparse, both spatially temporally. To augment missing information, generally RGB guidance leveraged due its high...
Large liquid crystal display (LCD) technology is being widely used in every corner of our modern life, ranging from personal laptops to flat-panel televisions. Among all the components an LCD system, backlight panel dominant power consumer, irrespective lighting or class. In this paper, a saliency-based field-programmable gate array accelerator for revolutionary management proposed that allows dynamic modulation different zones panel. This hardware accelerator-based system capable processing...
The ability to model and simulate guided wave propagation provides an insight into the development of robust inspection systems. This paper presents finite‐element method (FEM), as a powerful computational technique used in context ultrasonic nondestructive evaluation inaccessible plate like structures. A commercially available FE package is for this study. Corrosion type surface thinning with depths from 20% 50% are fabricated steel plate. series computer runs made using "Lamb type" waves...
Vision and video applications are becoming pervasive in mobile embedded systems. Consumer wearable devices require capabilities for real-time analytics prolonged battery lifetimes, which is further driving the need innovative system designs with low-power, reliability high performance. Further, increasing resolution of image sensors these systems places an demand on both memory storage as well computational power. Such stringent requirements have given rise to accelerator-rich architectures...