Jukka Heikkonen

Jatin Chaudhary, Ivan Jambor, Hannu Aronen et al.

Background: The reproducibility of machine-learning models in prostate cancer detection across different MRI vendors remains a significant challenge. Methods: This study investigates Support Vector Machines (SVM) and Random Forest (RF) models trained on radiomic features extracted from T2-weighted MRI images using Pyradiomics and MRCradiomics libraries. Feature selection was performed using the maximum relevance minimum redundancy (MRMR) technique. We aimed to enhance clinical decision support through multimodal learning and feature fusion. Results: Our SVM model, utilizing combined features from Pyradiomics and MRCradiomics, achieved an AUC of 0.74 on the Multi-Improd dataset (Siemens scanner) but decreased to 0.60 on the Philips test set. The RF model showed similar trends, with notable robustness for models using Pyradiomics features alone (AUC of 0.78 on Philips). Conclusions: These findings demonstrate the potential of multimodal feature integration to improve the robustness and generalizability of machine-learning models for clinical decision support in prostate cancer detection. This study marks a significant step towards developing reliable AI-driven diagnostic tools that maintain efficacy across various imaging platforms.

Sahar Salimpour, Jorge Peña-Queralta, Diego Paez-Granados et al.

Unprecedented agility and dexterous manipulation have been demonstrated with controllers based on deep reinforcement learning (RL), with a significant impact on legged and humanoid robots. Modern tooling and simulation platforms, such as NVIDIA Isaac Sim, have been enabling such advances. This article focuses on demonstrating the applications of Isaac in local planning and obstacle avoidance as one of the most fundamental ways in which a mobile robot interacts with its environments. Although there is extensive research on proprioception-based RL policies, the article highlights less standardized and reproducible approaches to exteroception. At the same time, the article aims to provide a base framework for end-to-end local navigation policies and how a custom robot can be trained in such simulation environment. We benchmark end-to-end policies with the state-of-the-art Nav2, navigation stack in Robot Operating System (ROS). We also cover the sim-to-real transfer process by demonstrating zero-shot transferability of policies trained in the Isaac simulator to real-world robots. This is further evidenced by the tests with different simulated robots, which show the generalization of the learned policy. Finally, the benchmarks demonstrate comparable performance to Nav2, opening the door to quick deployment of state-of-the-art end-to-end local planners for custom robot platforms, but importantly furthering the possibilities by expanding the state and action spaces or task definitions for more complex missions. Overall, with this article we introduce the most important steps, and aspects to consider, in deploying RL policies for local path planning and obstacle avoidance with Isaac Sim training, Gazebo testing, and ROS 2 for real-time inference in real robots. The code is available at https://github.com/sahars93/RL-Navigation.

Suyash Gaurav, Jukka Heikkonen, Jatin Chaudhary

As AI systems evolve into distributed ecosystems with autonomous execution, asynchronous reasoning, and multi-agent coordination, the absence of scalable, decoupled governance poses a structural risk. Existing oversight mechanisms are reactive, brittle, and embedded within agent architectures, making them non-auditable and hard to generalize across heterogeneous deployments. We introduce Governance-as-a-Service (GaaS): a modular, policy-driven enforcement layer that regulates agent outputs at runtime without altering model internals or requiring agent cooperation. GaaS employs declarative rules and a Trust Factor mechanism that scores agents based on compliance and severity-weighted violations. It enables coercive, normative, and adaptive interventions, supporting graduated enforcement and dynamic trust modulation. To evaluate GaaS, we conduct three simulation regimes with open-source models (LLaMA3, Qwen3, DeepSeek-R1) across content generation and financial decision-making. In the baseline, agents act without governance; in the second, GaaS enforces policies; in the third, adversarial agents probe robustness. All actions are intercepted, evaluated, and logged for analysis. Results show that GaaS reliably blocks or redirects high-risk behaviors while preserving throughput. Trust scores track rule adherence, isolating and penalizing untrustworthy components in multi-agent systems. By positioning governance as a runtime service akin to compute or storage, GaaS establishes infrastructure-level alignment for interoperable agent ecosystems. It does not teach agents ethics; it enforces them.

Suyash Gaurav, Muhammad Farhan Humayun, Jukka Heikkonen et al.

The evolution of Vision Transformers has led to their widespread adaptation to different domains. Despite large-scale success, there remain significant challenges including their reliance on extensive computational and memory resources for pre-training on huge datasets as well as difficulties in task-specific transfer learning. These limitations coupled with energy inefficiencies mainly arise due to the computation-intensive self-attention mechanism. To address these issues, we propose a novel Super-Pixel Based Patch Pooling (SPPP) technique that generates context-aware, semantically rich, patch embeddings to effectively reduce the architectural complexity and improve efficiency. Additionally, we introduce the Light Latent Attention (LLA) module in our pipeline by integrating latent tokens into the attention mechanism allowing cross-attention operations to significantly reduce the time and space complexity of the attention module. By leveraging the data-intuitive patch embeddings coupled with dynamic positional encodings, our approach adaptively modulates the cross-attention process to focus on informative regions while maintaining the global semantic structure. This targeted attention improves training efficiency and accelerates convergence. Notably, the SPPP module is lightweight and can be easily integrated into existing transformer architectures. Extensive experiments demonstrate that our proposed architecture provides significant improvements in terms of computational efficiency while achieving comparable results with the state-of-the-art approaches, highlighting its potential for energy-efficient transformers suitable for edge deployment. (The code is available on our GitHub repository: https://github.com/zser092/Focused-Attention-ViT).

Suyash Gaurav, Jukka Heikkonen, Jatin Chaudhary

Continual learning systems face the dual challenge of preventing catastrophic forgetting while maintaining energy efficiency, particularly in resource-constrained environments. This paper introduces Pathway-based Progressive Inference (PaPI), a novel theoretical framework that addresses these challenges through a mathematically rigorous approach to pathway selection and adaptation. We formulate continual learning as an energy-constrained optimization problem and provide formal convergence guarantees for our pathway routing mechanisms. Our theoretical analysis demonstrates that PaPI achieves an $\mathcal{O}(K)$ improvement in the stability-plasticity trade-off compared to monolithic architectures, where $K$ is the number of pathways. We derive tight bounds on forgetting rates using Fisher Information Matrix analysis and prove that PaPI's energy consumption scales with the number of active parameters rather than the total model size. Comparative theoretical analysis shows that PaPI provides stronger guarantees against catastrophic forgetting than Elastic Weight Consolidation (EWC) while maintaining better energy efficiency than both EWC and Gradient Episodic Memory (GEM). Our experimental validation confirms these theoretical advantages across multiple benchmarks, demonstrating PaPI's effectiveness for continual learning in energy-constrained settings. Our codes are available at https://github.com/zser092/PAPI_FILES.

Jatin Chaudhary, Dipak Nidhi, Jukka Heikkonen et al.

The objective of this paper is to enhance the optimization process for neural networks by developing a dynamic learning rate algorithm that effectively integrates exponential decay and advanced anti-overfitting strategies. Our primary contribution is the establishment of a theoretical framework where we demonstrate that the optimization landscape, under the influence of our algorithm, exhibits unique stability characteristics defined by Lyapunov stability principles. Specifically, we prove that the superlevel sets of the loss function, as influenced by our adaptive learning rate, are always connected, ensuring consistent training dynamics. Furthermore, we establish the "equiconnectedness" property of these superlevel sets, which maintains uniform stability across varying training conditions and epochs. This paper contributes to the theoretical understanding of dynamic learning rate mechanisms in neural networks and also pave the way for the development of more efficient and reliable neural optimization techniques. This study intends to formalize and validate the equiconnectedness of loss function as superlevel sets in the context of neural network training, opening newer avenues for future research in adaptive machine learning algorithms. We leverage previous theoretical discoveries to propose training mechanisms that can effectively handle complex and high-dimensional data landscapes, particularly in applications requiring high precision and reliability.

Arun Kunwar, Dibakar Raj Pant, Jukka Heikkonen et al.

The Convolutional Neural Network (CNN) has shown impressive performance in image classification because of its strong learning capabilities. However, it demands a substantial and balanced dataset for effective training. Otherwise, networks frequently exhibit over fitting and struggle to generalize to new examples. Publicly available dataset of fundus images of ocular disease is insufficient to train any classification model to achieve satisfactory accuracy. So, we propose Generative Adversarial Network(GAN) based data generation technique to synthesize dataset for training CNN based classification model and later use original disease containing ocular images to test the model. During testing the model classification accuracy with the original ocular image, the model achieves an accuracy rate of 78.6% for myopia, 88.6% for glaucoma, and 84.6% for cataract, with an overall classification accuracy of 84.6%.

Yu Xianjia, Jorge Peña Queralta, Jukka Heikkonen et al.

Autonomous systems are becoming inherently ubiquitous with the advancements of computing and communication solutions enabling low-latency offloading and real-time collaboration of distributed devices. Decentralized technologies with blockchain and distributed ledger technologies (DLTs) are playing a key role. At the same time, advances in deep learning (DL) have significantly raised the degree of autonomy and level of intelligence of robotic and autonomous systems. While these technological revolutions were taking place, raising concerns in terms of data security and end-user privacy has become an inescapable research consideration. Federated learning (FL) is a promising solution to privacy-preserving DL at the edge, with an inherently distributed nature by learning on isolated data islands and communicating only model updates. However, FL by itself does not provide the levels of security and robustness required by today's standards in distributed autonomous systems. This survey covers applications of FL to autonomous robots, analyzes the role of DLT and FL for these systems, and introduces the key background concepts and considerations in current research.

Yu Xianjia, Li Qingqing, Jorge Pena Queralta et al.

Unmanned aerial vehicles (UAVs) are becoming largely ubiquitous with an increasing demand for aerial data. Accurate navigation and localization, required for precise data collection in many industrial applications, often relies on RTK GNSS. These systems, able of centimeter-level accuracy, require a setup and calibration process and are relatively expensive. This paper addresses the problem of accurate positioning and navigation of UAVs through cooperative localization. Inexpensive ultra-wideband (UWB) transceivers installed on both the UAV and a support ground robot enable centimeter-level relative positioning. With fast deployment and wide setup flexibility, the proposed system is able to accommodate different environments and can also be utilized in GNSS-denied environments. Through extensive simulations and test fields, we evaluate the accuracy of the system and compare it to GNSS in urban environments where multipath transmission degrades accuracy. For completeness, we include visual-inertial odometry in the experiments and compare the performance with the UWB-based cooperative localization.

Xianjia Yu, Qingqing Li, Jorge Peña Queralta et al.

Ultra-wideband (UWB) technology is a mature technology that contested other wireless technologies in the advent of the IoT but did not achieve the same levels of widespread adoption. In recent years, however, with its potential as a wireless ranging and localization solution, it has regained momentum. Within the robotics field, UWB positioning systems are being increasingly adopted for localizing autonomous ground or aerial robots. In the Industrial IoT (IIoT) domain, its potential for ad-hoc networking and simultaneous positioning is also being explored. This survey overviews the state-of-the-art in UWB networking and localization for robotic and autonomous systems. We also cover novel techniques focusing on more scalable systems, collaborative approaches to localization, ad-hoc networking, and solutions involving machine learning to improve accuracy. This is, to the best of our knowledge, the first survey to put together the robotics and IIoT perspectives and to emphasize novel ranging and positioning modalities. We complete the survey with a discussion on current trends and open research problems.

Paavo Nevalainen, Parisa Movahedi, Jorge Peña Queralta et al.

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.

Sherif A. S. Mohamed, Jawad N. Yasin, Mohammad-Hashem Haghbayan et al.

Event cameras, i.e., the Dynamic and Active-pixel Vision Sensor (DAVIS) ones, capture the intensity changes in the scene and generates a stream of events in an asynchronous fashion. The output rate of such cameras can reach up to 10 million events per second in high dynamic environments. DAVIS cameras use novel vision sensors that mimic human eyes. Their attractive attributes, such as high output rate, High Dynamic Range (HDR), and high pixel bandwidth, make them an ideal solution for applications that require high-frequency tracking. Moreover, applications that operate in challenging lighting scenarios can exploit the high HDR of event cameras, i.e., 140 dB compared to 60 dB of traditional cameras. In this paper, a novel asynchronous corner tracking method is proposed that uses both events and intensity images captured by a DAVIS camera. The Harris algorithm is used to extract features, i.e., frame-corners from keyframes, i.e., intensity images. Afterward, a matching algorithm is used to extract event-corners from the stream of events. Events are solely used to perform asynchronous tracking until the next keyframe is captured. Neighboring events, within a window size of 5x5 pixels around the event-corner, are used to calculate the velocity and direction of extracted event-corners by fitting the 2D planar using a randomized Hough transform algorithm. Experimental evaluation showed that our approach is able to update the location of the extracted corners up to 100 times during the blind time of traditional cameras, i.e., between two consecutive intensity images.

Jawad N. Yasin, Sherif A. S. Mohamed, Mohammad-hashem Haghbayan et al.

Moving towards autonomy, unmanned vehicles rely heavily on state-of-the-art collision avoidance systems (CAS). However, the detection of obstacles especially during night-time is still a challenging task since the lighting conditions are not sufficient for traditional cameras to function properly. Therefore, we exploit the powerful attributes of event-based cameras to perform obstacle detection in low lighting conditions. Event cameras trigger events asynchronously at high output temporal rate with high dynamic range of up to 120 $dB$. The algorithm filters background activity noise and extracts objects using robust Hough transform technique. The depth of each detected object is computed by triangulating 2D features extracted utilising LC-Harris. Finally, asynchronous adaptive collision avoidance (AACA) algorithm is applied for effective avoidance. Qualitative evaluation is compared using event-camera and traditional camera.

Sherif A. S. Mohamed, Jawad N. Yasin, Mohammad-hashem Haghbayan et al.

Event-based cameras are vision devices that transmit only brightness changes with low latency and ultra-low power consumption. Such characteristics make event-based cameras attractive in the field of localization and object tracking in resource-constrained systems. Since the number of generated events in such cameras is huge, the selection and filtering of the incoming events are beneficial from both increasing the accuracy of the features and reducing the computational load. In this paper, we present an algorithm to detect asynchronous corners from a stream of events in real-time on embedded systems. The algorithm is called the Three Layer Filtering-Harris or TLF-Harris algorithm. The algorithm is based on an events' filtering strategy whose purpose is 1) to increase the accuracy by deliberately eliminating some incoming events, i.e., noise, and 2) to improve the real-time performance of the system, i.e., preserving a constant throughput in terms of input events per second, by discarding unnecessary events with a limited accuracy loss. An approximation of the Harris algorithm, in turn, is used to exploit its high-quality detection capability with a low-complexity implementation to enable seamless real-time performance on embedded computing platforms. The proposed algorithm is capable of selecting the best corner candidate among neighbors and achieves an average execution time savings of 59 % compared with the conventional Harris score. Moreover, our approach outperforms the competing methods, such as eFAST, eHarris, and FA-Harris, in terms of real-time performance, and surpasses Arc* in terms of accuracy.

Jawad N. Yasin, Sherif A. S. Mohamed, Mohammad-Hashem Haghbayan et al.

This work focuses on the formation reshaping in an optimized manner in autonomous swarm of drones. Here, the two main problems are: 1) how to break and reshape the initial formation in an optimal manner, and 2) how to do such reformation while minimizing the overall deviation of the drones and the overall time, i.e., without slowing down. To address the first problem, we introduce a set of routines for the drones/agents to follow while reshaping to a secondary formation shape. And the second problem is resolved by utilizing the temperature function reduction technique, originally used in the point set registration process. The goal is to be able to dynamically reform the shape of multi-agent based swarm in near-optimal manner while going through narrow openings between, for instance obstacles, and then bringing the agents back to their original shape after passing through the narrow passage using point set registration technique.

Jonne Pohjankukka, Sakari Tuominen, Jukka Heikkonen

In large-area forest inventories a trade-off between the amount of data to be sampled and the costs of collecting the data is necessary. It is not always possible to have a very large data sample when dealing with sampling-based inventories. It is therefore necessary to optimize the sampling design in order to achieve optimal population parameter estimation. On the contrary, the availability of remote sensing (RS) data correlated with the forest inventory variables is usually much higher. The combination of RS and the sampled field measurement data is often used for improving the forest inventory parameter estimation. In addition, it is also reasonable to study the utilization of RS data in inventory sampling, which can further improve the estimation of forest variables. In this study, we propose a data sampling method based on Bayesian optimization which uses RS data in forest inventory sample selection. The presented method applies the learned functional relationship between the RS and inventory data in new sampling decisions. We evaluate our method by conducting simulated sampling experiments with both synthetic data and measured data from the Aland region in Finland. The proposed method is benchmarked against two baseline methods: simple random sampling and the local pivotal method. The results of the simulated experiments show the best results in terms of MSE values for the proposed method when the functional relationship between RS and inventory data is correctly learned from the available training data.

Jonne Pohjankukka, Tapio Pahikkala, Paavo Nevalainen et al.

In machine learning one often assumes the data are independent when evaluating model performance. However, this rarely holds in practise. Geographic information data sets are an example where the data points have stronger dependencies among each other the closer they are geographically. This phenomenon known as spatial autocorrelation (SAC) causes the standard cross validation (CV) methods to produce optimistically biased prediction performance estimates for spatial models, which can result in increased costs and accidents in practical applications. To overcome this problem we propose a modified version of the CV method called spatial k-fold cross validation (SKCV), which provides a useful estimate for model prediction performance without optimistic bias due to SAC. We test SKCV with three real world cases involving open natural data showing that the estimates produced by the ordinary CV are up to 40% more optimistic than those of SKCV. Both regression and classification cases are considered in our experiments. In addition, we will show how the SKCV method can be applied as a criterion for selecting data sampling density for new research area.

Qingqing Li, Paavo Nevalainen, Jorge Peña Queralta et al.

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 observations are matched to a general 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 of 2100\,m of harvester path was recorded by an industrial harvester with a 3D laser scanner and a geolocation unit fixed to the frame. Our experiments show a 12\,cm s.t.d. in the location accuracy and with real-time data processing for speeds not exceeding 0.5\,m/s. The accuracy and speed limit is realistic during forest operations.

Markus Viljanen, Antti Airola, Jukka Heikkonen et al.

Maximizing product use is a central goal of many businesses, which makes retention and monetization two central analytics metrics in games. Player retention may refer to various duration variables quantifying product use: total playtime or session playtime are popular research targets, and active playtime is well-suited for subscription games. Such research often has the goal of increasing player retention or conversely decreasing player churn. Survival analysis is a framework of powerful tools well suited for retention type data. This paper contributes new methods to game analytics on how playtime can be analyzed using survival analysis without covariates. Survival and hazard estimates provide both a visual and an analytic interpretation of the playtime phenomena as a funnel type nonparametric estimate. Metrics based on the survival curve can be used to aggregate this playtime information into a single statistic. Comparison of survival curves between cohorts provides a scientific AB-test. All these methods work on censored data and enable computation of confidence intervals. This is especially important in time and sample limited data which occurs during game development. Throughout this paper, we illustrate the application of these methods to real world game development problems on the Hipster Sheep mobile game.