scholarly journals Deep Learning-Based Target Tracking and Classification for Low Quality Videos Using Coded Aperture Cameras

Sensors ◽  
2019 ◽  
Vol 19 (17) ◽  
pp. 3702 ◽  
Author(s):  
Chiman Kwan ◽  
Bryan Chou ◽  
Jonathan Yang ◽  
Akshay Rangamani ◽  
Trac Tran ◽  
...  
Keyword(s):  
Deep Learning ◽  
Power Consumption ◽  
Compressive Sensing ◽  
Target Tracking ◽  
Low Power Consumption ◽  
Coded Aperture ◽  
Learning Approach ◽  
Residual Network ◽  
Practical Applications ◽  
Measurement Domain

Compressive sensing has seen many applications in recent years. One type of compressive sensing device is the Pixel-wise Code Exposure (PCE) camera, which has low power consumption and individual control of pixel exposure time. In order to use PCE cameras for practical applications, a time consuming and lossy process is needed to reconstruct the original frames. In this paper, we present a deep learning approach that directly performs target tracking and classification in the compressive measurement domain without any frame reconstruction. In particular, we propose to apply You Only Look Once (YOLO) to detect and track targets in the frames and we propose to apply Residual Network (ResNet) for classification. Extensive simulations using low quality optical and mid-wave infrared (MWIR) videos in the SENSIAC database demonstrated the efficacy of our proposed approach.

scholarly journals Detection and Confirmation of Multiple Human Targets Using Pixel-Wise Code Aperture Measurements

2020 ◽  
Vol 6 (6) ◽  
pp. 40 ◽  
Author(s):  
Chiman Kwan ◽  
David Gribben ◽  
Akshay Rangamani ◽  
Trac Tran ◽  
Jack Zhang ◽  
...  
Keyword(s):  
Deep Learning ◽  
Data Storage ◽  
Target Detection ◽  
Vehicle Detection ◽  
Residual Network ◽  
Reconstruction Process ◽  
Recent Approach ◽  
Measurement Domain ◽  
Multiple Frames

Compressive video measurements can save bandwidth and data storage. However, conventional approaches to target detection require the compressive measurements to be reconstructed before any detectors are applied. This is not only time consuming but also may lose information in the reconstruction process. In this paper, we summarized the application of a recent approach to vehicle detection and classification directly in the compressive measurement domain to human targets. The raw videos were collected using a pixel-wise code exposure (PCE) camera, which condensed multiple frames into one frame. A combination of two deep learning-based algorithms (you only look once (YOLO) and residual network (ResNet)) was used for detection and confirmation. Optical and mid-wave infrared (MWIR) videos from a well-known database (SENSIAC) were used in our experiments. Extensive experiments demonstrated that the proposed framework was feasible for target detection up to 1500 m, but target confirmation needs more research.

Improved Baugh Wooley Multiplier Using Cadence Register Transfer Level Logic for Power Consumption

2017 ◽  
Vol 14 (1) ◽  
pp. 277-283
Author(s):  
V Rajmohan ◽  
O. Uma Maheswari
Keyword(s):  
Power Consumption ◽  
Low Power ◽  
Vlsi Design ◽  
Register Transfer Level ◽  
Low Power Consumption ◽  
Power Efficient ◽  
Practical Applications ◽  
Efficient Process ◽  
Array Multiplier ◽  
Register Transfer

In modern days of VLSI design, speedy operations and low-power consumption is a key requirement for any circuits. When it comes to multipliers, the power efficient multiplier plays an important role. The main aim of this work is to develop the system with faster and less power multiplier for an efficient process by using Baugh-Wooley multipliers. The optimized Baugh-Wooley multiplier consumes least power, area and produces less delay. The proposed architecture is 193× times faster than Conventional array multiplier in the practical applications and 213× times faster than a conventional Baugh-Wooley multiplier. The Improved Baugh-Wooley multiplier consumes the power of 09.02 mW and area of 52426 μm2.

Manipulating and trapping light with photonic crystals from fundamental studies to practical applications

2016 ◽  
Vol 4 (47) ◽  
pp. 11032-11049 ◽  
Author(s):  
Eiichi Kuramochi
Keyword(s):  
Photonic Crystals ◽  
Power Consumption ◽  
Low Power ◽  
Cavity Qed ◽  
Photonic Devices ◽  
Low Power Consumption ◽  
Practical Applications ◽  
Recent Advances

This review summarizes recent advances in trapping and extracting light, cavity-QED studies, and low power consumption photonic devices by photonic crystals and nanostructures.

A Portable and High-Resolution System for Sleep Apnea Monitoring

2013 ◽  
Vol 753-755 ◽  
pp. 2369-2373
Author(s):  
Yu Xuan Hu ◽  
Yi Hu ◽  
Shu Ming Ye ◽  
Xiao Xiang Zheng
Keyword(s):  
Sleep Apnea ◽  
High Resolution ◽  
Power Consumption ◽  
Low Power ◽  
Monitoring System ◽  
Sleep Apnea Syndrome ◽  
Low Power Consumption ◽  
Practical Applications ◽  
Obstructive Sleep ◽  
Apnea Monitoring

As a major indicator of Obstructive Sleep Apnea Syndrome (OSAS) in clinical diagnosis, the monitoring of sleep apnea plays an important role in medical treatments of modern society. This paper proposes a portable sleep apnea monitoring system, which is of high-precision and low-power consumption, and capable of performing the long-term monitoring of OSAS patients multiple physiological parameters in clinical treatments. In the system, the AC modulated detection is adopted, and low amplification ratios are utilized in forestage and a high-resolution AD converter is designed in post-stages. Thus, it is able to acquire, analyze, and process physiological signals in real-time. In addition, ultralow-power chips are used in control system to save the power consumption. The experimental results show that our monitoring system has the strengths of high stability, low-power consumption (peak current90mA), and strong anti-interference ability, which demonstrates the potential in practical applications.

scholarly journals Development of a camera trap for perching dragonflies: a new tool for freshwater environmental assessment

PeerJ ◽  
2020 ◽  
Vol 8 ◽  
pp. e9681
Author(s):  
Akira Yoshioka ◽  
Akira Shimizu ◽  
Hiroyuki Oguma ◽  
Nao Kumada ◽  
Keita Fukasawa ◽  
...  
Keyword(s):  
Power Consumption ◽  
Low Power ◽  
Low Cost ◽  
Time Lapse ◽  
Camera Trap ◽  
Environmental Indicators ◽  
Low Power Consumption ◽  
Detection Accuracy ◽  
Practical Applications ◽  
Detection Frequency

Although dragonflies are excellent environmental indicators for monitoring terrestrial water ecosystems, automatic monitoring techniques using digital tools are limited. We designed a novel camera trapping system with an original dragonfly detector based on the hypothesis that perching dragonflies can be automatically detected using inexpensive and energy-saving photosensors built in a perch-like structure. A trial version of the camera trap was developed and evaluated in a case study targeting red dragonflies (Sympetrum spp.) in Japan. During an approximately 2-month period, the detector successfully detected Sympetrum dragonflies while using extremely low power consumption (less than 5 mW). Furthermore, a short-term field experiment using time-lapse cameras for validation at three locations indicated that the detection accuracy was sufficient for practical applications. The frequency of false positive detection ranged from 17 to 51 over an approximately 2-day period. The detection sensitivities were 0.67 and 1.0 at two locations, where a time-lapse camera confirmed that Sympetrum dragonflies perched on the trap more than once. However, the correspondence between the detection frequency by the camera trap and the abundance of Sympetrum dragonflies determined by field observations conducted in parallel was low when the dragonfly density was relatively high. Despite the potential for improvements in our camera trap and its application to the quantitative monitoring of dragonflies, the low cost and low power consumption of the detector make it a promising tool.

scholarly journals Deteksi Penyakit Covid-19 Berdasarkan Citra X-Ray Menggunakan Deep Residual Network

2020 ◽  
Vol 8 (2) ◽  
pp. 443
Author(s):  
YULI SUN HARIYANI ◽  
SUGONDO HADIYOSO ◽  
THOMHERT SUPRAPTO SIADARI
Keyword(s):  
Deep Learning ◽  
Clinical Diagnosis ◽  
Learning Approach ◽  
Residual Network ◽  
X Ray ◽  
Global Pandemic ◽  
X Ray Imaging ◽  
Clinical Diagnoses ◽  
Chest X Ray ◽  
Supporting Tool

ABSTRAKPenyakit Coronavirus-2019 atau Covid-19 telah menjadi pandemi global dan menjadi masalah utama yang harus segera dikendalikan. Salah satu cara yang dapat dilakukan adalah memutus rantai penyebaran virus tersebut dengan melakukan deteksi dan melalukan karantina. Pencitraan X-Ray dapat dijadikan alternatif dalam mempelajari Covid-19. X-Ray dianggap mampu menggambarkan kondisi paru-paru pada pasien Covid-19 dan dapat menjadi alat bantu diagnosa klinis. Pada penelitian ini, kami mengusulkan pendekatan deep learning berbasis residual deep network untuk deteksi Covid-19 melalui citra chest X-Ray. Evaluasi yang dilakukan untuk mengetahui performa metode yang diusulkan berupa precision, recall, F1, dan accuracy. Hasil eksperimen menunjukkan bahwa usulan metode ini memberikan precision, recall, F1 dan accuracy masing-masing 0,98, 0,95, 0,97 dan 99%. Pada masa mendatang, studi ini diharapkan dapat divalidasi dan kemudian digunakan untuk melengkapi diagnosa klinis oleh dokter.Kata kunci: Coronavirus-2019, Covid-19, chest X-Ray, deep learning, residual network ABSTRACTCoronavirus-2019 or Covid-19 disease has become a global pandemic and is a major problem that must be stopped immediately. One of the ways that can be done to stop its spreading is to break the spreading chain of the virus by detecting and doing quarantine. X-Ray imaging can be used as an alternative in detecting Covid-19. X-Ray is considered able to describe the condition of the lungs for Covid-19 suspected patients and can be a supporting tool for clinical diagnosis. In this study, we propose a residual based deep learning approach for Covid-19 detection using chest X-Ray images. Evaluation is carried out to determine the performance of the proposed method in the form of precision, recall, F1 and accuracy. Experiments results show that our proposed method provides precision, recall, F1 and accuracy respectively 0.98, 0.95, 0.97 and 99%. In the future, this study is expected to be validated and then used to support clinical diagnoses by doctors.Keywords: Coronavirus-2019, Covid-19, chest X-Ray, deep learning, residual network

Sign in / Sign up

Export Citation Format

Share Document