Kaspars Ozols

Kaspars Ozols
Elektronikas un datorzinātņu institūts · Signal Processing Lab.

Dr.sc.ing.

About

27
Publications
15,062
Reads
How we measure 'reads'
A 'read' is counted each time someone views a publication summary (such as the title, abstract, and list of authors), clicks on a figure, or views or downloads the full-text. Learn more
206
Citations
Citations since 2017
20 Research Items
195 Citations
2017201820192020202120222023020406080100120
2017201820192020202120222023020406080100120
2017201820192020202120222023020406080100120
2017201820192020202120222023020406080100120
Introduction
My research interests includes Wireless Sensor Networks (WSN), audio signal processing, biometric, brain computer interfaces, EEG signal data acquistion and processing, non-uniform sampling, wireless communications, signal processing, smart sensor development etc. Apart from science, I'm strongly involved in writing many project proposals (eg. FP7, Horizon 2020). I have gained expertise both as coordinator and partner.
Education
September 2011 - June 2017
Riga Technical University
Field of study
  • Electronics

Publications

Publications (27)
Conference Paper
Full-text available
In this paper a method of encoding the signals by using Amplitude Adaptive Asynchronous Sigma-Delta modulator (AA-ASDM) scheme without an additional envelope encoding of the signal is proposed. According to AA-ASDM, the time-varying envelope function of the input signal is used in the feedback loop to reduce the switching rate of the output trigger...
Conference Paper
Full-text available
The paper presents an improved version of Asynchronous Sigma-Delta Modulator (ASDM) in order to reduce its power consumption for time-encoding of analog signals. The proposed method is called Amplitude Adaptive ASDM (AA-ASDM) since time-varying envelope of the signal is used in the feedback loop of ASDM to reduce the switching rate of the output tr...
Conference Paper
Full-text available
This paper describes a multichannel mobile EEG data acquisition system that consists of on-head sensors with built in electroencephalogram (EEG) signal amplifier, asynchronous sigma-delta modulator (ASDM) for analog to digital conversion and 434MHz On-Off keying (OOK) wireless data transmitter. A prototype circuit has been designed and fabricated i...
Conference Paper
Full-text available
In this paper an investigation is carried out on how to decrease the length of the FIR filter which is used for equalization of the loudspeaker distortions. It is shown that instead of one high-order filter, two FIR filters of lower order can be used by considering that an audio signal is split into two separate frequency bands by a digital crossov...
Conference Paper
Full-text available
Wireless sensor network (WSN) testbed design challenges can be divided into three main sub-categories: architectural, hardware and software. By considering all these challenges, an optimal solution was found and EDI TestBed created. In this paper, a method, which solves software related problems in large scale WSN testbed, such as how to distribute...
Article
The availability of data is an important aspect of any research as it determines the likelihood of the study's commencement, completion, and success. The Internet of Things and Wireless Sensor Networks technologies have been attracting a huge amount of researchers for more than two decades, without having a consolidated or unified source, identifyi...
Article
Full-text available
Semantic segmentation of an incoming visual stream from cameras is an essential part of the perception system of self-driving cars. State-of-the-art results in semantic segmentation have been achieved with deep neural networks (DNNs), yet training them requires large datasets, which are difficult and costly to acquire and time-consuming to label. A...
Article
Full-text available
Model understanding is critical in many domains, particularly those involved in high-stakes decisions, e.g., medicine, criminal justice, and autonomous driving. Explainable AI (XAI) methods are essential for working with black-box models such as convolutional neural networks. This paper evaluates the traffic sign classifier of the Deep Neural Netwo...
Article
Full-text available
Automated Driving Systems (ADSs) commend a substantial reduction of human-caused road accidents while simultaneously lowering emissions, mitigating congestion, decreasing energy consumption and increasing overall productivity. However, achieving higher SAE levels of driving automation and complying with ISO26262 C and D Automotive Safety Integrity...
Preprint
Model understanding is critical in many domains, particularly those involved in high-stakes decisions, i.e., medicine, criminal justice, and autonomous driving. Explainable AI (XAI) methods are essential for working with black-box models such as Convolutional Neural Networks. This paper evaluates the traffic sign classifier of Deep Neural Network (...
Article
Full-text available
Deep neural networks (DNNs) have achieved state-of-the-art results in a broad range of tasks, in particular the ones dealing with the perceptual data. However, full-scale application of DNNs in safety-critical areas is hindered by their black box-like nature, which makes their inner workings nontransparent. As a response to the black box problem, t...
Article
Full-text available
Modern machines strive to run at limit performance and dependability while their operational area and size are getting restricted. To achieve those objectives, often swift integration of custom-made subsystems is required, either actuators, sensors, electronic, or SW modules. Such a diverse suite of elements needs specific approaches and tools for...
Conference Paper
Full-text available
Autonomous driving is disrupting the automotive industry as we know it today. For this, fail-operational behavior is essential in the sense, plan, and act stages of the automation chain in order to handle safety-critical situations on its own, which currently is not reached with state-of-the-art approaches. The European ECSEL research project PRYST...
Article
Full-text available
Smart manufacturing and smart factories depend on automation and robotics, whereas human–robot collaboration (HRC) contributes to increasing the effectiveness and productivity of today’s and future factories. Industrial robots especially in HRC settings can be hazardous if safety is not addressed properly. In this review, we look at the collaborati...
Article
Full-text available
A large number of different low-power wireless network technologies exist, including IEEE 802.15.4, Bluetooth Low Energy, multiple protocol standards of WiFi (IEEE 802.11), Near Field Communication (NFC), and LoRa. Given this number of competing technologies, the selection of the best one for a new project is not trivial. This paper aims to help In...
Article
Full-text available
Artificial Neural Networks (ANNs) have become an accepted approach for a wide range of challenges. Meanwhile, the advancement of chip manufacturing processes is approaching saturation which calls for new computing solutions. This work presents a novel approach of an FPGA-based accelerator development for fully connected feed-forward neural networks...
Conference Paper
Full-text available
Autonomous driving has the potential to disruptively change the automotive industry as we know it today. For this, fail-operational behavior is essential in the sense, plan, and act stages of the automation chain in order to handle safety-critical situations by its own, which currently is not reached with state-of-the-art approaches.The European EC...
Article
Full-text available
The paper describes the elements of the developed MATLAB Simulink library for building the models of Bluetooth Low Energy (BLE) wireless sensor networks to simulate the communication between BLE devices in the presence of interference and channel noise. Various parameters can be configured for the devices including their 2D positions to take into a...
Article
Full-text available
This paper describes an implementation of reception and real-time decoding of Asynchronous Sigma-Delta modulator (ASDM) encoded and wirelessly (On-off keying) transmitted signals. By using Fast reconstruction algorithm it is possible to reconstruct the original signal ≈ 25 times faster than by Classical reconstruction algorithm, if the signal lengh...
Conference Paper
Full-text available
In this paper a design of FIR decimation filters with low group delay (less than 1 millisecond) is investigated. Given an input sampling rate of 192 kHz, linear-phase and minimum-phase FIR filters are obtained with decimation factors 16 and 32, respectively. Simulation results and the frequency responses of the minimum-phase filters are provided.

Network

Cited By

Projects

Projects (3)
Project
The wearable sensor platform proposed in CONVERGENCE is centred on energy efficient wearable proof-of-concepts at system level exploiting data analytics developed in a context driven approach (in contrast with more traditional research where the device level research and the data analytics are carried out on separate path, rarely converging).