Paavo Nevalainen - Academia.edu (original) (raw)

Papers by Paavo Nevalainen

AHFE International

Organizational service learn-leadership design for adapting and predicting machine learning-based... more Organizational service learn-leadership design for adapting and predicting machine learning-based sentiments of sociotechnical systems is being addressed in segmenting textual-producing agents in classes. In the past, there have been numerous demonstrations in different language models (LMs) and (naıve) Bayesian Networks (BN) that can classify textual knowledge origin for different classes based on decisive binary trees from the future prediction aspect of how public text collection and processing can be approached, converging the root causes of events. An example is how communication influence and affect the end-user. Within service providers and industry, the progress of processing communication relies on formal clinical and informal non-practices. The LM is based on handcrafted division on machine learning (ML) approaches representing the subset of AI and can be used as an orthogonal policy-as-a-target leadership tool in customer or political discussions. The classifiers which us...

Remote Sensing

Abnormal behavior detection is currently receiving much attention because of the availability of ... more Abnormal behavior detection is currently receiving much attention because of the availability of marine equipment and data allowing maritime agents to track vessels. One of the most popular tools for developing an efficient anomaly detection system is the Automatic Identification System (AIS). The aim of this paper is to explore the performance of existing well-known clustering methods for detecting the two most dangerous abnormal behaviors based on the AIS. The methods include K-means, Density-Based Spatial Clustering of Applications with Noise (DBSCAN), Affinity Propagation (AP), and the Gaussian Mixtures Model (GMM). In order to evaluate the performance of the clustering methods, we also used the AIS data of vessels, which were collected through the Finnish transport agency from the whole Baltic Sea for three months. Although most existing studies focus on ocean route recognition, deviations from regulated ocean routes, or irregular speed, we focused on dark ships or those sets o...

2023 26th International Conference on Information Fusion (FUSION)

The originality of this thesis has been checked in accordance with the University of Turku qualit... more The originality of this thesis has been checked in accordance with the University of Turku quality assurance system using the Turnitin OriginalityCheck service. Asiasanat: laiva, päästö, tuulikalibraatio, dispersiomalli UNIVERSITY OF TURKU Department of Future Technologies ALI LEINO: On remote matching of ships to pollution Master's thesis, 77 p., 4 app. p. Computer Science September 2019 Air pollution from shipping emissions is a significant hazard for humans and the environment. International Maritime Organization (IMO) has introduced limits on the used fuel and exhaust emissions from ships. A fixed measurement station can be used to remotely measure exhaust gases from passing ships to indicate their compliance with regulations. Automatic operation of a station requires that it can pair a measurement to a ship. Data from a measurement station in Turku archipelago was used to collect gas concentrations, atmospheric data and Automatic Identification System (AIS) measurements from passing ships. Ship location information and class (A or B) were used from the AIS data. Data for the years 2017 and 2018 was used, with the latter used as a test set. A prediction model is designed to predict a concentration time series for a ship's route in the current atmospheric conditions. The model is based on the Gaussian puff approach with traditional Pasquill stability classes to estimate dispersion parameters. The time of maximum concentration and its value is extracted from the predictions for each passing ship. A separate peak detection model is used to extract peaks from measured gas concentrations that are significant and may originate from passing ships. Every predicted peak is matched to temporally closest observed peak to arrive at a list of possible matches of ships to their measured pollution. A new family of classification metrics is introduced that take into account the probability of matches happening at random. A modification of the F 1-score, F 1t-score was used as a performance metric. One set of parameters of the model control which predicted and observed peaks are used for matching. These were optimized using Random Search (RS). We show that without the random performance adjustment the model can't be optimized to results that represent ships matched to their pollution. Further the effect of including Class B ships is shown to always reduce the performance of the model. A novel approach is used to learn a wind calibration function to correct for the effect of nearby obstacles. With the parameters optimized by RS locked, the parameters of the wind calibration are optimized using Stochastic Gradient Descent (SGD). This process is continued with another round of RS and SGD and is hence called iterative RS+SGD. Results of the wind calibration were validated using both visual comparison and by calculating the root-mean-square error (RMSE) with a nearby weather station maintained by Finnish Meteorological Institute (FMI). The results show that the method learned a correction that better represents the true wind conditions than without the correction. Iterative RS+SGD improved the F 1t-score while lowering the amount of random matches.

Remote Sensing, 2020

Autonomous harvesting and transportation is a long-term goal of the forest industry. One of the m... more Autonomous harvesting and transportation is a long-term goal of the forest industry. One of the main challenges is the accurate localization of both vehicles and trees in a forest. Forests are unstructured environments where it is difficult to find a group of significant landmarks for current fast feature-based place recognition algorithms. This paper proposes a novel approach where local point clouds are matched to a global tree map using the Delaunay triangularization as the representation format. Instead of point cloud based matching methods, we utilize a topology-based method. First, tree trunk positions are registered at a prior run done by a forest harvester. Second, the resulting map is Delaunay triangularized. Third, a local submap of the autonomous robot is registered, triangularized and matched using triangular similarity maximization to estimate the position of the robot. We test our method on a dataset accumulated from a forestry site at Lieksa, Finland. A total length o...

Proceedings of the 6th International Conference on Pattern Recognition Applications and Methods, 2017

Curvature spectrum is a useful feature in surface classification but is difficult to apply to cas... more Curvature spectrum is a useful feature in surface classification but is difficult to apply to cases with high noise typical e.g. to natural resource point clouds. We propose two methods to estimate the mean and the Gaussian curvature with filtering properties specific to triangulated surfaces. Methods completely filter a highest shape mode away but leave single vertical pikes only partially dampened. Also an elaborate computation of nodal dual areas used by the Laplace-Beltrami mean curvature can be avoided. All computation is based on triangular setting, and a weighted summation procedure using projected tip angles sums up the vertex values. A simplified principal curvature direction definition is given to avoid computation of the full second fundamental form. Qualitative evaluation is based on numerical experiments over two synthetical examples and a prostata tumor example. Results indicate the proposed methods are more robust to presence of noise than other four reference formulations.

2015 International Conference on Image Processing Theory, Tools and Applications (IPTA), 2015

Presence of ground surface stones is one indicator of economically important landmass deposits in... more Presence of ground surface stones is one indicator of economically important landmass deposits in the Arctic. The other indicator is a geomorphological category of the area. This work shows that ground stoniness can be automatically predicted with practical accuracy. Northern forests have less biomass and foliage, thus direct analysis of stoniness is possible from airborne laser scanning (ALS) data. A test set of 88 polygons covering 3.3 km2 was human-classified and a method was developed to perform the stoniness classification over this set. The local curvature of the surface is approximated directly from the point cloud data without generating the Digital Terrain Model (DTM). The method performs well with area under curve AUC = 0.85 from Leave-Pair-Out cross-validation, and is rather insensitive to missing data, moderate forest cover and double-scanned areas.

ABSTRACT The structure of a business simulation environment (Simo-1.00) is presented. General goa... more ABSTRACT The structure of a business simulation environment (Simo-1.00) is presented. General goals of the software project are presented. Some technical aspects in business activity modeling are discussed. It is ar gued that business simulation is another end of the spectrum of computer-aided networking. The other end is vir- tual organizations. One goal of the Simo-1.00 is to pro vide a platform where var- ious virtual organization models can be tested before actual changes in the management level are made and before investment to the modern communica- tions software is done.

Tutkimusraportti - Geologian Tutkimuskeskus

The project investigated the possibilities of publicly available spatial data in mapping and pred... more The project investigated the possibilities of publicly available spatial data in mapping and predicting of geospatial phenomena with economical importance. The aim was to develop practical applications based merely on open data from the databases of Finnish Meteorological Institute (FMI), Geological Survey of Finland (GTK), Finnish Forest Research Institute (Metla; since 1st Jan., 2015, Natural Resources Institute Finland, Luke) and National Land Survey of Finland (NLS). Geographic Information Systems (GIS), Remote Sensing (RS) and Machine Learning (ML) techniques were applied in developing various applications. The most promising applications were: i) Hydrological Operations and Prediction Model, HOPS, ii) Mapping of mass-flow aggregate deposits for infrastructure construction, iii) Quick response mapping of forest floods, and, iv) Mapping of drainage networks. The HOPS is already an operational application, while other applications still need further validation to become operation...

2022 International Conference on Electrical, Computer and Energy Technologies (ICECET)

2018 IEEE 27th International Symposium on Industrial Electronics (ISIE), 2018

In the recent years, big data and learning analytics have been emerging as fast-growing research ... more In the recent years, big data and learning analytics have been emerging as fast-growing research fields. The application of these emerging research areas is gradually addressing the contemporary challenges of school and university education. Tracing out the information regarding students' misconceptions and dropping-out probabilities from the courses at the right instant of time, development of detectors of a range of educational importance and achieving the highest level of quality in the higher education are becoming more challenging. Moreover, providing well timed and the best suitable solutions to the students at-risk are even more strenuous. In this concept paper, we aim to address these contemporary challenges of school and the university education and their probable solutions by utilizing our research experiences of automated assessment, immediate feedback, learning analytics and the IT technologies. Solving such problems by knowing the history of students' activities, submissions, and the performances data is possible. The identification of students' misconceptions during the learning process, examining behavioral patterns and significant trends by efficiently aggregating and correlating the massive data, improving the state-of-the-art skills in creative thinking and innovation, and detecting the drop-outs on-time are highlighted in this article. We are aiming at extracting such knowledge so that adaptive and personalized learning will become a part of the current education system. Not only the available algorithm of supervised learning methods such as support vector machine, neural network, decision trees, discriminant analysis, and nearest neighborhood method but also new engineering and distillation of relevant data features can be carried out to solve these educational challenges.

2018 Baltic Geodetic Congress (BGC Geomatics), 2018

This paper proposes a new approach for the detection of tree locations around forest machines pro... more This paper proposes a new approach for the detection of tree locations around forest machines producing a situational model based on on-site terrestrial LiDAR data collected during harvesting operation. A triangularized ground model is used to planarize the point cloud in order to simplify the tree detection. The planarized ground makes the vertical cutting of the point cloud systematical. Tree stem lines detected from individual trees at individual scan views are used to guide the final alignment into global coordinates. The setup is numerically efficient and does not rely on any positioning and orientation system (POS) based e.g. on an inertial measurement unit (IMU) or global navigation satellite system (GNSS) or wheel rotation counter on the autonomous vehicle.

Multiorganizational technological projects are commonplace today. The ways of cooperation between... more Multiorganizational technological projects are commonplace today. The ways of cooperation between organizations vary. A common basis for interorganizational project planning is needed. Simo-2 developed at the Prosit 9 Virtual delivery project is such a tool. It is a distributed business simulation environment, which has product model based approach for power plant deliveries. It brings a unified and concrete representation of intended project structure. This paper presents the model structure of Simo-2 and describes how the software can be used in preliminary project planning. INTRODUCTION Power plant deliveries are multiorganizational technological projects. The structure of the consortium and way of cooperation between organizations vary from project to project, and this brings added difficulty in managing these one-of-a-kind product delivery projects. A common basis for interorganizational project planning is needed. Simo-2 is such a tool. It is a distributed business simulation ...

2018 17th IEEE International Conference on Machine Learning and Applications (ICMLA), 2018

In this paper, we propose an effective object detection framework based on proposal fusion of mul... more In this paper, we propose an effective object detection framework based on proposal fusion of multiple sensors such as infrared camera, RGB cameras, radar and LiDAR. Our framework first applies the Selective Search (SS) method on RGB image data to extract possible candidate proposals which likely contain the objects of interest. Then it uses the information from other sensors in order to reduce the number of generated proposals by SS and find more dense proposals. Finally, the class of objects within the final proposals are identified by Convolutional Neural Network (CNN). Experimental results on real dataset demonstrate that our framework can precisely detect meaningful object regions using a smaller number of proposals than other object proposals methods. Further, our framework can achieve reliable object detection and classification results in maritime environments.

2018 21st International Conference on Intelligent Transportation Systems (ITSC), 2018

Robust real-time object detection and tracking are challenging problems in autonomous transportat... more Robust real-time object detection and tracking are challenging problems in autonomous transportation systems due to operation of algorithms in inherently uncertain and dynamic environments and rapid movement of objects. Therefore, tracking and detection algorithms must cooperate with each other to achieve smooth tracking of detected objects that later can be used by the navigation system. In this paper, we first present an efficient multi-sensor fusion approach based on the probabilistic data association method in order to achieve accurate object detection and tracking results. The proposed approach fuses the detection results obtained independently from four main sensors: radar, LiDAR, RGB camera and infrared camera. It generates object region proposals based on the fused detection result. Then, a Convolutional Neural Network (CNN) approach is used to identify the object categories within these regions. The CNN is trained on a real dataset from different ferry driving scenarios. The experimental results of tracking and classification on real datasets show that the proposed approach provides reliable object detection and classification results in maritime environments.

2019 IEEE International Conference on Artificial Intelligence Circuits and Systems (AICAS), 2019

In recent years, Artificial Intelligence (AI) has been widely deployed in a variety of business s... more In recent years, Artificial Intelligence (AI) has been widely deployed in a variety of business sectors and industries, yielding numbers of revolutionary applications and services that are primarily driven by high-performance computation and storage facilities in the cloud. On the other hand, embedding intelligence into edge devices is highly demanded by emerging applications such as autonomous systems, human-machine interactions, and the Internet of Things (IoT). In these applications, it is advantageous to process data near or at the source of data to improve energy & spectrum efficiency and security, and decrease latency. Although the computation capability of edge devices has increased tremendously during the past decade, it is still challenging to perform sophisticated AI algorithms in these resource-constrained edge devices, which calls for not only low-power chips for energy efficient processing at the edge but also a system-level framework to distribute resources and tasks along the edge-cloud continuum. In this overview, we summarize dedicated edge hardware for machine learning from embedded applications to sub-mW "always-on" IoT nodes. Recent advances of circuits and systems incorporating joint design of architectures and algorithms will be reviewed. Fog computing paradigm that enables processing at the edge while still offering the possibility to interact with the cloud will be covered, with focus on opportunities and challenges of exploiting fog computing in AI as a bridge between the edge device and the cloud.

New Developments and Environmental Applications of Drones, 2021

Vehicles with prolonged autonomous missions have to maintain environment awareness by simultaneou... more Vehicles with prolonged autonomous missions have to maintain environment awareness by simultaneous localization and mapping (SLAM). Closed loop correction is substituted by interpolation in rigid body transformation space in order to systematically reduce the accumulated error over different scales. The computation is divided to an edge computed lightweight SLAM and iterative corrections in the cloud environment. Tree locations in the forest environment are sent via a potentially limited communication bandwidths. Data from a real forest site is used in the verification of the proposed algorithm. The algorithm adds new iterative closest point (ICP) cases to the initial SLAM and measures the resulting map quality by the mean of the root mean squared error (RMSE) of individual tree clusters. Adding 4 % more match cases yields the mean RMSE 0.15 m on a large site with 180 m odometric distance.

Geomorphology, 2020

This is a PDF file of an article that has undergone enhancements after acceptance, such as the ad... more This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.