Abstract Traditional explainability methods use 2D visualization techniques such as saliency maps. However, when applied to video data, subject and camera motion produce unstable, flickering maps that cannot be temporally aggregated in meaningful ways. This limitation is critical in medical and veterinary settings, which demand biologically grounded explanations intuitive for experts. One such domain is equine pain assessment. Horses are particularly challenging in the context of pain as they are known to hide pain signals in human presence. This leads to increased interest in automation of equine pain recognition which is explainable, i.e., highlighting informative facial and postural cues. However, current explainability approaches are inadequate for providing temporally consistent and clinically interpretable explanations. To address this gap, we present SHIC-XE, an explainability framework that leverages dense 2D-3D correspondence methods to address fundamental viewpoint-dependency limitations for dynamic scenes. Building upon the existing SHIC correspondence framework, our approach projects neural network attention onto canonical 3D surface prototypes, enabling anatomically-consistent, pose-invariant explanations across temporal sequences. Our primary contribution is establishing quantitative metrics for spatially-consistent explainability evaluation, transforming subjective visualization assessment into statistically analyzable measurements. Unlike existing approaches requiring n separate classifiers for n anatomical regions, our unified correspondence mapping maintains constant architectural complexity while achieving comparable classification performance. We provide a preliminary validation of the framework on equine pain recognition as a challenging testbed, achieving video-level F1 scores of 0.67, 0.80, and 0.70 across three datasets under leave-one-subject-out validation. Our systematic evaluation against expert pain ratings by facial region, using a validated equine pain scale, reveals significant correlations between model attention and expert pain scoring, (ears: r = 0.30, $${\varvec{p}} \varvec{<} {\textbf {0.001}}$$ p < 0.001 ; cheek muscles: r = 0.25, $${\varvec{p}} \varvec{<} {\textbf {0.01}}$$ p < 0.01 ; eyes: r = −0.19, $${\textbf {p}} \varvec{<} {\textbf {0.05}}$$ p < 0.05 ) providing the first quantitative validation of explainability methods against domain expertise in this context. This finding challenges widespread literature claims of clinical relevance based solely on qualitative saliency inspection, establishing a methodological framework for rigorous explainability validation. The correspondence-based approach generalizes beyond the demonstrated application to any domain requiring spatial consistency between 2D observations and 3D structural models, providing a foundational methodology for next-generation explainable AI systems.

Abstract Neural Style Transfer (NST) is a technique that applies the visual characteristics of one image to another while preserving the structural content. Traditionally used for artistic transformations, NST has more recently been adopted in new fields, such as domain adaptation and data augmentation. This study explores a novel application of the technique as a data augmentation strategy for landmark detection. To validate the approach as a feasibility study for a complex cat facial landmark detection task, we make two main contributions. First, we demonstrate that applying style transfer to cropped facial images, rather than full-body images, improves the structural consistency of generated images, which is necessary for the landmark detection task. Secondly, we suggest a Supervised Neural Style Transfer (SNST) technique specifically adapted for the landmark detection task, which outperforms traditional image augmentation methods and unsupervised NST. Our findings propose SNST as an effective augmentation strategy for improving the performance of landmark detection models in the animal domain and potentially beyond.

Abstract The field of animal affective computing is rapidly emerging, and analysis of facial expressions is a crucial aspect. One of the most significant challenges that researchers in the field currently face is the scarcity of high-quality, comprehensive datasets that allow the development of models for facial expressions analysis. One of the possible approaches is the utilisation of facial landmarks, which has been shown for humans and animals. In this paper we present a novel dataset of cat facial images annotated with bounding boxes and 48 facial landmarks grounded in cat facial anatomy. We also introduce a landmark detection convolution neural network-based model which uses a magnifying ensemble method. Our model shows excellent performance on cat faces and is generalizable to human and other animals facial landmark detection.

Abstract Advances in animal motion tracking and pose recognition have been a game changer in the study of animal behavior. Recently, an increasing number of works go ‘deeper’ than tracking, and address automated recognition of animals’ internal states such as emotions and pain with the aim of improving animal welfare, making this a timely moment for a systematization of the field. This paper provides a comprehensive survey of computer vision-based research on recognition of pain and emotional states in animals, addressing both facial and bodily behavior analysis. We summarize the efforts that have been presented so far within this topic—classifying them across different dimensions, highlight challenges and research gaps, and provide best practice recommendations for advancing the field, and some future directions for research.

In many cases, software artifacts share similarity across projects and development teams. However, often this similarity is only partially reflected on the level of design and implementation, and therefore the possibilities for its detection are limited in current variability analysis, clone detection, and application search approaches. In this paper, we propose a method for identification and comparison of similarly behaving software. The method, supported by a prototype tool, analyzes the behavioral similarity of object-oriented code artifacts based on shallow (behavior interface) and deep (behavior transformation) descriptions of the exhibited operations. It further recommends on suitable mechanisms inspired by the notion of polymorphism in order to guide and support current and future reuse. The approach was evaluated on two data-sets, obtained following two different scenarios: clone-and-own and independent development by different teams.

Abstract We report on a mixed-method, comparative study investigating whether there is a difference between privacy concerns expressed about pet wearables as opposed to human wearables – and more importantly,why. We extracted the privacy concerns found in product reviews (N=8,038) of pet wearables (activity, location, and dual-function trackers), contrasting the (lack of) concerns and misuse to a curated set of reviews for similar human-oriented wearables (N=20,431). Our findings indicate that, while overall very few privacy concerns are expressed in product reviews, for pet wearables they are expressed even less, even though consumers use these devices in a manner which impacts both personal and bystander privacy. An additional survey of pet owners (N=201) eliciting what factors would cause them to not purchase (or stop using) pet wearables indicated comparably few privacy concerns, strengthening the representativeness of our findings. A thematic analysis reveals that the lack of privacy concerns may be explained by, among other factors, emotional drivers to purchase the device, and prioritization of (desired) functionality to support those emotional drivers over privacy requirements. Moreover, we found that pet wearables are used in different ways than originally intended, which raise novel privacy implications to be dealt with. We propose that in order to move towards more privacy-conscious use of pet wearables, a combination of understanding consumer rationale and behavior as well as ensuring data protection legislation is adequate to real-world use is needed.

As software becomes more complex, reusing and integrating artifacts from existing projects that may be taken from open or organization-proprietary repositories is becoming an increasingly important practice. This practice requires an in-depth understanding of the projects to be reused and particularly their common and variable features and their non-functional requirements. Different approaches have been suggested to analyze similarity and variability of different kinds of artifacts (mainly, requirements and code), e.g., clone detection and feature mining. These approaches, however, mainly address functional aspects of the software artifacts, while mostly neglecting aspects dictated by non-functional requirements. The recent progress with the General Data Protection Regulation (GDPR) highlights the importance of handling privacy concerns in software development. However, existing approaches do not directly refer to privacy challenges in software reuse. In this paper we propose integrating these two lines of research and introduce a privacy-aware software reuse approach. Particularly, we suggest to extend VarMeR – Variability Mechanisms Recommender – which analyzes software similarity based on exhibited behaviors and recommends on polymorphism-inspired reuse mechanisms, with privacy awareness considerations. These considerations are reflected in “privacy levels” of the reused artifacts.

The use of lightweight formal methods (LFM) for the development of industrial applications has become a major trend. Although the term "lightweight formal methods" has been used for over ten years now, there seems to be no common agreement on what "lightweight" actually means, and different communities apply the term in all kinds of ways. In this paper, we explore the recent trends in the use of LFM, and establish our opinion that cost-effectiveness is the driving force to deploy LFM. Further, we propose a simple framework that should help to classify different LFM approaches and to estimate which of them are most cost-effective for a certain software engineering project. We demonstrate our framework using some examples.

This paper proposes a systematic framework for applying the Physics of Notations (PoN), a theory for the design of cognitively effective visual notations. The PoN consists of nine principles, but not all principles lend themselves equally to a clear and unambiguous operationalization. As a result, many visual notations designed according to the PoN apply it in different ways. The proposed framework guides what information is required of a reported PoN application to ensure that the application of each principle is verifiable. The framework utilizes an evidence-driven design rationale model to structure information needed to assess principles requiring user involvement or cognitive theories. This approach aims to reduce ambiguity in some of the principles by making design choices explicit, and highlighting the level of evidence presented to support it. We demonstrate the proposed framework in a showcase of a recently published visual notation which has been designed with the PoN in mind.

With the advance of modern technologies, computer-based systems for animals are gaining popularity. In particular, there is an explosion of products and gadgets for pets: wellness monitoring applications (e.g., FitBark and PetPace), automatic food dispensers, cognitive enrichment apps, and many more. Furthermore, the discipline of Animal-Computer Interaction has emerged, focusing on a user-centric development of technologies for animals, making them stakeholders in the development process. Animal-centric technologies have already been developed to support activities of rescue and assistance dogs, to provide environmental and cognitive enrichment for animals in captivity, and to support conservation and animal behavior research. Going beyond human stakeholders poses new exciting challenges for requirement engineering and can be used to significantly expand its boundaries under broader theoretical and methodological frameworks. This paper highlights these challenges and proposes a research agenda for developing methodologies for requirement elicitation and analysis for a user-centric development of computerized systems for non-human users.

This paper presents AHR, a formal framework for combinatorial test design that is Agile, Human-centred and Refinement-oriented. The framework (i) allows us to reuse test plans developed for an abstract level at more concrete levels; (ii) has human-centric interface providing queries and alerts whenever the specified test plan is incomplete or invalid; (iii) involves analysis of the testing constraints within combinatorial testing.

The use of Formal Methods (FMs) offers rigour and precision, while reducing ambiguity and inconsistency. The major barriers hindering the adoption of FMs in industry are the problems of understandability, comprehensibility, and scalability. To solve the understandability problem, from one side, the readability of the method have to be increased, but from another side, an appropriate teaching and learning approach have to be introduced. This paper presents an overview of existing approaches on teaching of FMs and Logic, also discusses the common issues in teaching of this subjects.

In this paper we address particular aspects of integration of formal methods in large-scale industrial projects, namely collaborative aspects. We review recent works addressing such aspects, identify some current trends and discuss directions for further research.

New visual languages and extensions of existing notations are increasingly introduced for various purposes. There are many frameworks and theories that aid designers in creating cognitively effective visual notations. They usually provide general guidelines, which are not easily operationalized without extensive user involvement. However, often such user involvement is difficult to achieve, especially in scientific settings where finding enough participants with the needed background knowledge is far from trivial. In this paper we propose the idea of creating a searchable marketplace of visual elements, in which designers could share and exchange elements of visual notations. A key feature of such marketplace would be the ability to certify such elements via procedures grounded in empirical research. We discuss a proposed structure of the marketplace, and the challenge that the need for certification poses to its design.

The ability to capture different levels of abstraction in a system model is especially important for remote integration, testing/verification, and manufacturing of cyber-physical systems (CPSs). However, the complexity of modelling and testing of CPSs makes these processes extremely prone to human error. In this paper we present our ongoing work on introducing human-centred considerations into modelling and testing of CPSs, which allow for agile iterative refinement processes of different levels of abstraction when errors are discovered or missing information is completed.

Coherence Management refers to all efforts one needs to invest, in order to ensure that information shown in, and implied by a representation of requirements makes sense as a whole, is coherent. Coherence Management is an umbrella term we use to cover, and more importantly, stimulate research on relationships between identification, measurement, and action on phenomena which reflect tensions between information in requirements representations. Such tensions exist between information which is, for example, logically inconsistent, or stakeholders disagree on, or signals tradeoffs (meaning that improvement on some requirements, for instance, necessarily means some quantifiable (or not) deterioration of others). These tensions are an important topic of research in Requirements Engineering, and various methods have been proposed for the identification, measurement, and action on logical inconsistency in requirements models, on negotiating disagreements, and on settling tradeoffs. Despite focusing on related phenomena, these methods are different and each come with their own specific definition of when a representation of requirements is incoherent and what to do about it. This makes it hard to compare existing methods, design new ones, and choose those to apply when doing RE. In this short communication we outline our research agenda for developing a unified formal framework for the systematization and classification of Coherence Management efforts in the context of RE, as well as exploring their compatibility.

Inconsistency is a major challenge in requirements engineering, commonly perceived as a problem that needs to be eliminated on sight. However, in practice maintaining consistency at all times is an intractable problem. Accordingly, recent paradigms for inconsistency management acknowledge that it is sometimes desirable to tolerate inconsistency, e.g. to allow distributed teamwork and prevent premature commitment to design decisions. However, a successful adoption of inconsistency managment paradigms in industry depends highly on the human factor: intolerant attitudes of practitioners toward inconsistency may pose significant barriers to a wider acceptance of these paradigms. A thorough analysis of cognitive factors is a key to overcoming these barriers. This paper reports on our preliminary empirical findings highlighting existing perceptions and attitudes of practitioners toward inconsistency, and propose dimensions for their classification. Based on these results, we outline a general research program for exploring cognitive factors in inconsistency management.

We develop a fully algorithmic approach to “taming” logics expressed Hilbert style, that is, reformulating them in terms of analytic sequent calculi and useful semantics. Our approach applies to Hilbert calculi extending the positive fragment of propositional classical logic with axioms of a certain general form that contain new unary connectives. Our work encompasses various results already obtained for specific logics. It can be applied to new logics, as well as to known logics for which an analytic calculus or a useful semantics has so far not been available. A Prolog implementation of the method is described.

In the paper Cut-free ordinary sequent calculi for logics having generalized finite-valued semantics by A. Avron, J. Ben-Naim, and B. Konikowska. (Logica Universalis, 1:41–69, 2006), a general method was developed for generating cut-free ordinary sequent calculi for logics that can be characterized by finite-valued semantics based on non-deterministic matrices (Nmatrices). In this paper, a substantial step towards automation of paraconsistent reasoning is made by applying that method to a certain crucial family of thousands of paraconsistent logics, all belonging to the class of C-systems. For that family, the method produces in a modular way uniform Gentzen-type rules corresponding to a variety of axioms considered in the literature.

This paper makes a substantial step towards automatization of Para consistent reasoning by providing a general method for a systematic and modular generation of cut-free calculi for thousands of Para consistent logics known as Logics of Formal (In)consistency. The method relies on the use of non-deterministic semantics for these logics.

Maximality is a desirable property of paraconsistent logics, motivated by the aspiration to tolerate inconsistencies, but at the same time retain as much as possible from classical logic. In this paper we introduce a new, strong notion of maximal paraconsistency, which is based on possible extensions of the consequence relation of a logic. We investigate this notion in the framework of finite-valued paraconsistent logics, and show that for every n > 2 there exists an extensive family of n-valued logics, each of which is maximally paraconsistent in our sense, is partial to classical logic, and is not equivalent to any k-valued logic with k <; n. On the other hand, we specify a natural condition that guarantees that a paraconsistent logic is contained in a logic in the class of three-valued paraconsistent logics, and show that all reasonably expressive logics in this class are maximal.

Propositional canonical Gentzen-type systems, introduced in 2001 by Avron and Lev, are systems which in addition to the standard axioms and structural rules have only logical rules in which exactly one occurrence of a connective is introduced and no other connective is mentioned. A constructive coherence criterion for the non-triviality of such systems was defined and it was shown that a system of this kind admits cut-elimination iff it is coherent. The semantics of such systems is provided using two-valued non-deterministic matrices (2Nmatrices). In 2005 Zamansky and Avron extended these results to systems with unary quantifiers of a very restricted form. In this paper we substantially extend the characterization of canonical systems to (n,k)-ary quantifiers, which bind k distinct variables and connect n formulas, and show that the coherence criterion remains constructive for such systems. Then we focus on the case of k∈{0,1} and for a canonical calculus G show that it is coherent precisely when it has a strongly characteristic 2Nmatrix, which in turn is equivalent to admitting strong cut-elimination.