Michael Riegler

Thu Nguyen, Rabindra Khadka, Nhan Phan et al.

For many use cases, combining information from different datasets can be of interest to improve a machine learning model's performance, especially when the number of samples from at least one of the datasets is small. However, a potential challenge in such cases is that the features from these datasets are not identical, even though there are some commonly shared features among the datasets. To tackle this challenge, we propose a novel framework called Combine datasets based on Imputation (ComImp). In addition, we propose a variant of ComImp that uses Principle Component Analysis (PCA), PCA-ComImp in order to reduce dimension before combining datasets. This is useful when the datasets have a large number of features that are not shared between them. Furthermore, our framework can also be utilized for data preprocessing by imputing missing data, i.e., filling in the missing entries while combining different datasets. To illustrate the power of the proposed methods and their potential usages, we conduct experiments for various tasks: regression, classification, and for different data types: tabular data, time series data, when the datasets to be combined have missing data. We also investigate how the devised methods can be used with transfer learning to provide even further model training improvement. Our results indicate that the proposed methods are somewhat similar to transfer learning in that the merge can significantly improve the accuracy of a prediction model on smaller datasets. In addition, the methods can boost performance by a significant margin when combining small datasets together and can provide extra improvement when being used with transfer learning.

Sidney Pontes-Filho, Kristoffer Olsen, Anis Yazidi et al.

In this work, we argue that the search for Artificial General Intelligence (AGI) should start from a much lower level than human-level intelligence. The circumstances of intelligent behavior in nature resulted from an organism interacting with its surrounding environment, which could change over time and exert pressure on the organism to allow for learning of new behaviors or environment models. Our hypothesis is that learning occurs through interpreting sensory feedback when an agent acts in an environment. For that to happen, a body and a reactive environment are needed. We evaluate a method to evolve a biologically-inspired artificial neural network that learns from environment reactions named Neuroevolution of Artificial General Intelligence (NAGI), a framework for low-level AGI. This method allows the evolutionary complexification of a randomly-initialized spiking neural network with adaptive synapses, which controls agents instantiated in mutable environments. Such a configuration allows us to benchmark the adaptivity and generality of the controllers. The chosen tasks in the mutable environments are food foraging, emulation of logic gates, and cart-pole balancing. The three tasks are successfully solved with rather small network topologies and therefore it opens up the possibility of experimenting with more complex tasks and scenarios where curriculum learning is beneficial.

Thu Nguyen, Tuan L. Vo, Pål Halvorsen et al.

Missing data is a common problem in practical data science settings. Various imputation methods have been developed to deal with missing data. However, even though the labels are available in the training data in many situations, the common practice of imputation usually only relies on the input and ignores the label. We propose Classification Based on MissForest Imputation (CBMI), a classification strategy that initializes the predicted test label with missing values and stacks the label with the input for imputation, allowing the label and the input to be imputed simultaneously. In addition, we propose the imputation using labels (IUL) algorithm, an imputation strategy that stacks the label into the input and illustrates how it can significantly improve the imputation quality. Experiments show that CBMI has classification accuracy when the test set contains missing data, especially for imbalanced data and categorical data. Moreover, for both the regression and classification, IUL consistently shows significantly better results than imputation based on only the input data.

Vladimir Monakhov, Vajira Thambawita, Pål Halvorsen et al.

The interest for video anomaly detection systems has gained traction for the past few years. The current approaches use deep learning to perform anomaly detection in videos, but this approach has multiple problems. For starters, deep learning in general has issues with noise, concept drift, explainability, and training data volumes. Additionally, anomaly detection in itself is a complex task and faces challenges such as unknowness, heterogeneity, and class imbalance. Anomaly detection using deep learning is therefore mainly constrained to generative models such as generative adversarial networks and autoencoders due to their unsupervised nature, but even they suffer from general deep learning issues and are hard to train properly. In this paper, we explore the capabilities of the Hierarchical Temporal Memory (HTM) algorithm to perform anomaly detection in videos, as it has favorable properties such as noise tolerance and online learning which combats concept drift. We introduce a novel version of HTM, namely, Grid HTM, which is an HTM-based architecture specifically for anomaly detection in complex videos such as surveillance footage.

Sushant Gautam, Michael A. Riegler, Pål Halvorsen

Medical Visual Question Answering (MedVQA) is a promising field for developing clinical decision support systems, yet progress is often limited by the available datasets, which can lack clinical complexity and visual diversity. To address these gaps, we introduce Kvasir-VQA-x1, a new, large-scale dataset for gastrointestinal (GI) endoscopy. Our work significantly expands upon the original Kvasir-VQA by incorporating 159,549 new question-answer pairs that are designed to test deeper clinical reasoning. We developed a systematic method using large language models to generate these questions, which are stratified by complexity to better assess a model's inference capabilities. To ensure our dataset prepares models for real-world clinical scenarios, we have also introduced a variety of visual augmentations that mimic common imaging artifacts. The dataset is structured to support two main evaluation tracks: one for standard VQA performance and another to test model robustness against these visual perturbations. By providing a more challenging and clinically relevant benchmark, Kvasir-VQA-x1 aims to accelerate the development of more reliable and effective multimodal AI systems for use in clinical settings. The dataset is fully accessible and adheres to FAIR data principles, making it a valuable resource for the wider research community. Code and data: https://github.com/Simula/Kvasir-VQA-x1 and https://huggingface.co/datasets/SimulaMet/Kvasir-VQA-x1

Sushant Gautam, Michael A. Riegler, Pål Halvorsen

Vision-language models (VLMs) enable open-ended visual question answering but remain prone to hallucinations. We present HEDGE, a unified framework for hallucination detection that combines controlled visual perturbations, semantic clustering, and robust uncertainty metrics. HEDGE integrates sampling, distortion synthesis, clustering (entailment- and embedding-based), and metric computation into a reproducible pipeline applicable across multimodal architectures. Evaluations on VQA-RAD and KvasirVQA-x1 with three representative VLMs (LLaVA-Med, Med-Gemma, Qwen2.5-VL) reveal clear architecture- and prompt-dependent trends. Hallucination detectability is highest for unified-fusion models with dense visual tokenization (Qwen2.5-VL) and lowest for architectures with restricted tokenization (Med-Gemma). Embedding-based clustering often yields stronger separation when applied directly to the generated answers, whereas NLI-based clustering remains advantageous for LLaVA-Med and for longer, sentence-level responses. Across configurations, the VASE metric consistently provides the most robust hallucination signal, especially when paired with embedding clustering and a moderate sampling budget (n ~ 10-15). Prompt design also matters: concise, label-style outputs offer clearer semantic structure than syntactically constrained one-sentence responses. By framing hallucination detection as a geometric robustness problem shaped jointly by sampling scale, prompt structure, model architecture, and clustering strategy, HEDGE provides a principled, compute-aware foundation for evaluating multimodal reliability. The hedge-bench PyPI library enables reproducible and extensible benchmarking, with full code and experimental resources available at https://github.com/Simula/HEDGE .

Konstantinos Kousias, Apostolos Pappas, Ozgu Alay et al.

Bandwidth forecasting in Mobile Broadband (MBB) networks is a challenging task, particularly when coupled with a degree of mobility. In this work, we introduce HINDSIGHT++, an open-source R-based framework for bandwidth forecasting experimentation in MBB networks with Long Short Term Memory (LSTM) networks. We instrument HINDSIGHT++ following an Automated Machine Learning (AutoML) paradigm to first, alleviate the burden of data preprocessing, and second, enhance performance related aspects. We primarily focus on bandwidth forecasting for Fifth Generation (5G) networks. In particular, we leverage 5Gophers, the first open-source attempt to measure network performance on operational 5G networks in the US. We further explore the LSTM performance boundaries on Fourth Generation (4G) commercial settings using NYU-METS, an open-source dataset comprising of hundreds of bandwidth traces spanning different mobility scenarios. Our study aims to investigate the impact of hyperparameter optimization on achieving state-of-the-art performance and beyond. Results highlight its significance under 5G scenarios showing an average Mean Absolute Error (MAE) decrease of near 30% when compared to prior state-of-the-art values. Due to its universal design, we argue that HINDSIGHT++ can serve as a handy software tool for a multitude of applications in other scientific fields.

Sushant Gautam, Mehdi Houshmand Sarkhoosh, Jan Held et al.

The application of Automatic Speech Recognition (ASR) technology in soccer offers numerous opportunities for sports analytics. Specifically, extracting audio commentaries with ASR provides valuable insights into the events of the game, and opens the door to several downstream applications such as automatic highlight generation. This paper presents SoccerNet-Echoes, an augmentation of the SoccerNet dataset with automatically generated transcriptions of audio commentaries from soccer game broadcasts, enhancing video content with rich layers of textual information derived from the game audio using ASR. These textual commentaries, generated using the Whisper model and translated with Google Translate, extend the usefulness of the SoccerNet dataset in diverse applications such as enhanced action spotting, automatic caption generation, and game summarization. By incorporating textual data alongside visual and auditory content, SoccerNet-Echoes aims to serve as a comprehensive resource for the development of algorithms specialized in capturing the dynamics of soccer games. We detail the methods involved in the curation of this dataset and the integration of ASR. We also highlight the implications of a multimodal approach in sports analytics, and how the enriched dataset can support diverse applications, thus broadening the scope of research and development in the field of sports analytics.

Alireza Nik, Michael A. Riegler, Pål Halvorsen

Decoding strategies significantly influence the quality and diversity of the generated text in Large Language Models (LLMs), yet their impact on computational resources, particularly GPU energy consumption, is insufficiently studied. This paper investigates the relationship between text generation decoding techniques and energy efficiency, focusing on the trade-off between generation quality and GPU energy usage across diverse tasks and decoding configurations. By benchmarking multiple strategies across various tasks, including Translation, Math Problem Solving, Coding, and Open-ended text generation, we reveal how selecting appropriate decoding techniques with their tuned hyperparameters affects text quality and has measurable implications for energy consumption. Our findings show that the choice of decoding strategy can greatly impact GPU energy usage, even when it has a minimal effect on output quality. Different strategies also involve trade-offs between quality and energy efficiency, and no single decoding method is best in all cases across every metric. To the best of our knowledge, this is one of the first studies to examine decoding strategies in LLMs from the perspective of energy consumption, providing useful insights for building energy-efficient applications without compromising text generation quality.

Nhan Phan, Thu Nguyen, Uyen Dang et al.

Principal Component Analysis (PCA) is a commonly used tool for dimension reduction and denoising. Therefore, it is also widely used on the data prior to training a neural network. However, this approach can complicate the explanation of eXplainable Artificial Intelligence (XAI) methods for the decision of the model. In this work, we analyze the potential issues with this approach and propose Principal Components-based Initialization (PCsInit), a strategy to incorporate PCA into the first layer of a neural network via initialization of the first layer in the network with the principal components, and its two variants PCsInit-Act and PCsInit-Sub. We will show that explanations using these strategies are more simple, direct and straightforward than using PCA prior to training a neural network on the principal components. We also show that the proposed techniques possess desirable theoretical properties. Moreover, as will be illustrated in the experiments, such training strategies can also allow further improvement of training via backpropagation compared to training neural networks on principal components.

Birk Torpmann-Hagen, Michael A. Riegler, Pål Halvorsen et al.

Deep neural networks are being utilized in a growing number of applications, both in production systems and for personal use. Network checkpoints are as a consequence often shared and distributed on various platforms to ease the development process. This work considers the threat of neural network stegomalware, where malware is embedded in neural network checkpoints at a negligible cost to network accuracy. This constitutes a significant security concern, but is nevertheless largely neglected by the deep learning practitioners and security specialists alike. We propose the first effective countermeasure to these attacks. In particular, we show that state-of-the-art neural network stegomalware can be efficiently and effectively neutralized through shuffling the column order of the weight- and bias-matrices, or equivalently the channel-order of convolutional layers. We show that this effectively corrupts payloads that have been embedded by state-of-the-art methods in neural network steganography at no cost to network accuracy, outperforming competing methods by a significant margin. We then discuss possible means by which to bypass this defense, additional defense methods, and advocate for continued research into the security of machine learning systems.

Oriana Presacan, Alireza Nik, Vajira Thambawita et al.

Large Language Models (LLMs) rely on various decoding strategies to generate text, and these choices can significantly affect output quality. In healthcare, where accuracy is critical, the impact of decoding strategies remains underexplored. We investigate this effect in five open-ended medical tasks, including translation, summarization, question answering, dialogue, and image captioning, evaluating 11 decoding strategies with medically specialized and general-purpose LLMs of different sizes. Our results show that deterministic strategies generally outperform stochastic ones: beam search achieves the highest scores, while η and top-k sampling perform worst. Slower decoding methods tend to yield better quality. Larger models achieve higher scores overall but have longer inference times and are no more robust to decoding. Surprisingly, while medical LLMs outperform general ones in two of the five tasks, statistical analysis shows no overall performance advantage and reveals greater sensitivity to decoding choice. We further compare multiple evaluation metrics and find that correlations vary by task, with MAUVE showing weak agreement with BERTScore and ROUGE, as well as greater sensitivity to the decoding strategy. These results highlight the need for careful selection of decoding methods in medical applications, as their influence can sometimes exceed that of model choice.

Tuan L. Vo, Thu Nguyen, Luis M. Lopez-Ramos et al.

Missing data is a prevalent issue that can significantly impair model performance and explainability. This paper briefly summarizes the development of the field of missing data with respect to Explainable Artificial Intelligence and experimentally investigates the effects of various imputation methods on SHAP (SHapley Additive exPlanations), a popular technique for explaining the output of complex machine learning models. Next, we compare different imputation strategies and assess their impact on feature importance and interaction as determined by Shapley values. Moreover, we also theoretically analyze the effects of missing values on Shapley values. Importantly, our findings reveal that the choice of imputation method can introduce biases that could lead to changes in the Shapley values, thereby affecting the explainability of the model. Moreover, we also show that a lower test prediction MSE (Mean Square Error) does not necessarily imply a lower MSE in Shapley values and vice versa. Also, while XGBoost (eXtreme Gradient Boosting) is a method that could handle missing data directly, using XGBoost directly on missing data can seriously affect explainability compared to imputing the data before training XGBoost. This study provides a comprehensive evaluation of imputation methods in the context of model explanations, offering practical guidance for selecting appropriate techniques based on dataset characteristics and analysis objectives. The results underscore the importance of considering imputation effects to ensure robust and reliable insights from machine learning models.

Syed Zohaib Hassan, Kashif Ahmad, Michael A. Riegler et al.

The Visual Sentiment Analysis task is being offered for the first time at MediaEval. The main purpose of the task is to predict the emotional response to images of natural disasters shared on social media. Disaster-related images are generally complex and often evoke an emotional response, making them an ideal use case of visual sentiment analysis. We believe being able to perform meaningful analysis of natural disaster-related data could be of great societal importance, and a joint effort in this regard can open several interesting directions for future research. The task is composed of three sub-tasks, each aiming to explore a different aspect of the challenge. In this paper, we provide a detailed overview of the task, the general motivation of the task, and an overview of the dataset and the metrics to be used for the evaluation of the proposed solutions.

Imanol Luengo, Maria Grammatikopoulou, Rahim Mohammadi et al.

Surgical scene segmentation is essential for anatomy and instrument localization which can be further used to assess tissue-instrument interactions during a surgical procedure. In 2017, the Challenge on Automatic Tool Annotation for cataRACT Surgery (CATARACTS) released 50 cataract surgery videos accompanied by instrument usage annotations. These annotations included frame-level instrument presence information. In 2020, we released pixel-wise semantic annotations for anatomy and instruments for 4670 images sampled from 25 videos of the CATARACTS training set. The 2020 CATARACTS Semantic Segmentation Challenge, which was a sub-challenge of the 2020 MICCAI Endoscopic Vision (EndoVis) Challenge, presented three sub-tasks to assess participating solutions on anatomical structure and instrument segmentation. Their performance was assessed on a hidden test set of 531 images from 10 videos of the CATARACTS test set.

Debesh Jha, Sharib Ali, Nikhil Kumar Tomar et al.

Minimally invasive surgery is a surgical intervention used to examine the organs inside the abdomen and has been widely used due to its effectiveness over open surgery. Due to the hardware improvements such as high definition cameras, this procedure has significantly improved and new software methods have demonstrated potential for computer-assisted procedures. However, there exists challenges and requirements to improve detection and tracking of the position of the instruments during these surgical procedures. To this end, we evaluate and compare some popular deep learning methods that can be explored for the automated segmentation of surgical instruments in laparoscopy, an important step towards tool tracking. Our experimental results exhibit that the Dual decoder attention network (DDANet) produces a superior result compared to other recent deep learning methods. DDANet yields a Dice coefficient of 0.8739 and mean intersection-over-union of 0.8183 for the Robust Medical Instrument Segmentation (ROBUST-MIS) Challenge 2019 dataset, at a real-time speed of 101.36 frames-per-second that is critical for such procedures.

Abhishek Srivastava, Debesh Jha, Sukalpa Chanda et al.

Methods based on convolutional neural networks have improved the performance of biomedical image segmentation. However, most of these methods cannot efficiently segment objects of variable sizes and train on small and biased datasets, which are common for biomedical use cases. While methods exist that incorporate multi-scale fusion approaches to address the challenges arising with variable sizes, they usually use complex models that are more suitable for general semantic segmentation problems. In this paper, we propose a novel architecture called Multi-Scale Residual Fusion Network (MSRF-Net), which is specially designed for medical image segmentation. The proposed MSRF-Net is able to exchange multi-scale features of varying receptive fields using a Dual-Scale Dense Fusion (DSDF) block. Our DSDF block can exchange information rigorously across two different resolution scales, and our MSRF sub-network uses multiple DSDF blocks in sequence to perform multi-scale fusion. This allows the preservation of resolution, improved information flow and propagation of both high- and low-level features to obtain accurate segmentation maps. The proposed MSRF-Net allows to capture object variabilities and provides improved results on different biomedical datasets. Extensive experiments on MSRF-Net demonstrate that the proposed method outperforms the cutting-edge medical image segmentation methods on four publicly available datasets. We achieve the dice coefficient of 0.9217, 0.9420, and 0.9224, 0.8824 on Kvasir-SEG, CVC-ClinicDB, 2018 Data Science Bowl dataset, and ISIC-2018 skin lesion segmentation challenge dataset respectively. We further conducted generalizability tests and achieved a dice coefficient of 0.7921 and 0.7575 on CVC-ClinicDB and Kvasir-SEG, respectively.

Syed Zohaib Hassan, Kashif Ahmad, Steven Hicks et al.

The increasing popularity of social networks and users' tendency towards sharing their feelings, expressions, and opinions in text, visual, and audio content, have opened new opportunities and challenges in sentiment analysis. While sentiment analysis of text streams has been widely explored in literature, sentiment analysis from images and videos is relatively new. This article focuses on visual sentiment analysis in a societal important domain, namely disaster analysis in social media. To this aim, we propose a deep visual sentiment analyzer for disaster related images, covering different aspects of visual sentiment analysis starting from data collection, annotation, model selection, implementation, and evaluations. For data annotation, and analyzing peoples' sentiments towards natural disasters and associated images in social media, a crowd-sourcing study has been conducted with a large number of participants worldwide. The crowd-sourcing study resulted in a large-scale benchmark dataset with four different sets of annotations, each aiming a separate task. The presented analysis and the associated dataset will provide a baseline/benchmark for future research in the domain. We believe the proposed system can contribute toward more livable communities by helping different stakeholders, such as news broadcasters, humanitarian organizations, as well as the general public.

Kashif Ahmad, Konstantin Pogorelov, Mohib Ullah et al.

In this paper we present our methods for the MediaEval 2019 Mul-timedia Satellite Task, which is aiming to extract complementaryinformation associated with adverse events from Social Media andsatellites. For the first challenge, we propose a framework jointly uti-lizing colour, object and scene-level information to predict whetherthe topic of an article containing an image is a flood event or not.Visual features are combined using early and late fusion techniquesachieving an average F1-score of82.63,82.40,81.40and76.77. Forthe multi-modal flood level estimation, we rely on both visualand textual information achieving an average F1-score of58.48and46.03, respectively. Finally, for the flooding detection in time-based satellite image sequences we used a combination of classicalcomputer-vision and machine learning approaches achieving anaverage F1-score of58.82%