Unsupervised industrial anomaly detection requires accurately identifying defects without labeled data. Traditional autoencoder-based methods often struggle with incomplete anomaly suppression and loss of fine details, as their single-pass decoding fails to effectively handle anomalies with varying severity and scale. We propose a recursive architecture for autoencoder (RcAE), which performs reconstruction iteratively to progressively suppress anomalies while refining normal structures. Unlike traditional single-pass models, this recursive design naturally produces a sequence of reconstructions, progressively exposing suppressed abnormal patterns. To leverage this reconstruction dynamics, we introduce a Cross Recursion Detection (CRD) module that tracks inconsistencies across recursion steps, enhancing detection of both subtle and large-scale anomalies. Additionally, we incorporate a Detail Preservation Network (DPN) to recover high-frequency textures typically lost during reconstruction. Extensive experiments demonstrate that our method significantly outperforms existing non-diffusion methods, and achieves performance on par with recent diffusion models with only 10% of their parameters and offering substantially faster inference. These results highlight the practicality and efficiency of our approach for real-world applications.

The interpretability of deep neural networks is crucial for understanding model decisions in various applications, including computer vision. In this paper, we propose AttEXplore+, a unified adversarial attribution framework built upon AttEXplore that connects transferable adversarial exploration with gradient-path attribution. Rather than treating transferable attacks merely as additional attack choices, AttEXplore+ reformulates them as gradient acquisition operators for constructing smoother and more structured decision-boundary exploration paths. We instantiate AttEXplore++ with validated operators such as MIG and GRA, and conduct extensive experiments on five models, including CNNs (Inception-v3, ResNet-50, VGG16) and vision transformers (MaxViT-T, ViT-B/16), using the ImageNet dataset. Our method achieves an average performance improvement of 7.57% over AttEXplore and 32.62% compared to other state-of-the-art interpretability algorithms. Using insertion and deletion scores as evaluation metrics, we show that adversarial transferability plays a vital role in enhancing attribution results. Furthermore, we explore the impact of randomness, perturbation rate, noise amplitude, and diversity probability on attribution performance, demonstrating that AttEXplore++ provides more stable and reliable explanations across various models. We release our code at: https://github.com/KxPlaug/ATTEXPLOREP .

The task of identifying multimodal image-text representations has garnered increasing attention, particularly with models such as CLIP (Contrastive Language-Image Pretraining), which demonstrate exceptional performance in learning complex associations between images and text. Despite these advancements, ensuring the interpretability of such models is paramount for their safe deployment in real-world applications, such as healthcare. While numerous interpretability methods have been developed for unimodal tasks, these approaches often fail to transfer effectively to multimodal contexts due to inherent differences in the representation structures. Bottleneck methods, well-established in information theory, have been applied to enhance CLIP's interpretability. However, they are often hindered by strong assumptions or intrinsic randomness. To overcome these challenges, we propose the Narrowing Information Bottleneck Theory, a novel framework that fundamentally redefines the traditional bottleneck approach. This theory is specifically designed to satisfy contemporary attribution axioms, providing a more robust and reliable solution for improving the interpretability of multimodal models. In our experiments, compared to state-of-the-art methods, our approach enhances image interpretability by an average of 9\%, text interpretability by an average of 58.83\%, and accelerates processing speed by 63.95\%. Our code is publicly accessible at https://github.com/LMBTough/NIB.

Out-of-distribution (OOD) detection is essential for enhancing the robustness and security of deep learning models in unknown and dynamic data environments. Gradient-based OOD detection methods, such as GAIA, analyse the explanation pattern representations of in-distribution (ID) and OOD samples by examining the sensitivity of model outputs w.r.t. model inputs, resulting in superior performance compared to traditional OOD detection methods. However, we argue that the non-zero gradient behaviors of OOD samples do not exhibit significant distinguishability, especially when ID samples are perturbed by random perturbations in high-dimensional spaces, which negatively impacts the accuracy of OOD detection. In this paper, we propose a novel OOD detection method called S & I based on layer Splitting and gradient Integration via Adversarial Gradient Attribution. Specifically, our approach involves splitting the model’s intermediate layers and iteratively updating adversarial examples layer-by-layer. We then integrate the attribution gradients from each intermediate layer along the attribution path from adversarial examples to the actual input, yielding true explanation pattern representations for both ID and OOD samples. Experiments demonstrate that our S & I algorithm achieves state-of-the-art results, with the average FPR95 of 29.05% (ResNet34)/38.61% (WRN40) and 37.31% (BiT-S) on the CIFAR100 and ImageNet benchmarks, respectively. Our code is available at: https://github.com/LMBTough/S-Ihttps://github.com/LMBTough/S-I

Reproducibility is indispensable for transferring explainable-AI algorithms from academic prototypes to production systems. This companion paper documents the artefacts, procedures, and outcomes that reproduce the empirical claims of ''Enhancing Model Interpretability with Local Attribution over Global Exploration'' (ACM MM 2024). We release a containerised archive containing source code, data-serialisation scripts, one-click executables, and a detailed README, all conforming to the ACM Multimedia reproducibility guidelines. The regenerated Insertion and Deletion scores deviate by only 2.2% on average. In addition, an exhaustive 10, 20, 30 3 grid-search over key hyper-parameters reveals a new configuration, (30, 20, 30), that improves the Insertion score of three convolutional backbones by 7.51% without additional code changes. These artefacts provide a rigorous, extensible foundation for future research on local attribution methods. Our code is available at: https://github.com/LMBTough/LA/

Imitation learning demonstrates remarkable performance in various domains. However, imitation learning is also constrained by many prerequisites. The research community has done intensive research to alleviate these constraints, such as introducing a stochastic policy to avoid unseen states, eliminating the need for action labels, and learning from the suboptimal demonstrations. Inspired by the natural reproduction process, we proposed a method called GenIL that integrates genetic operations with imitation learning. The involvement of the genetic operations improves the data efficiency by reproducing trajectories with various returns and assists the model in estimating more accurate and compact reward function parameters. We tested GenIL in both Atari and Mujoco domains, and the empirical result shows that it successfully outperforms the previous extrapolation methods in terms of extrapolation accuracy and overall policy performance when input data is limited. Code, data, and supplementary materials are publicly available at https://github.com/BryanZ666/GenIL.

Imitation learning (IL) aims to extract knowledge from human experts' demonstrations or artificially created agents to replicate their behaviors. It promotes interdisciplinary communication and real-world automation applications. However, the process of replicating behaviors still exhibits various problems, such as the performance is highly dependent on the demonstration quality, and most trained agents are limited to perform well in task-specific environments. In this survey, we provide an insightful review on IL. We first introduce the background knowledge from development history and preliminaries, followed by presenting different taxonomies within IL and key milestones of the field. We then detail challenges in learning strategies and present research opportunities with learning policy from suboptimal demonstration, voice instructions, and other associated optimization schemes.

Graph neural networks (GNNs) have proven their efficacy in a variety of real-world applications, but their underlying mechanisms remain a mystery. To address this challenge and enable reliable decision-making, many GNN explainers have been proposed in recent years. However, these methods often encounter limitations, including their dependence on specific instances, lack of generalizability to unseen graphs, producing potentially invalid explanations, and yielding inadequate fidelity. To overcome these limitations, we, in this paper, introduce the Auxiliary Classifier Generative Adversarial Network (ACGAN) into the field of GNN explanation and propose a new GNN explainer dubbed ACGAN-GNNExplainer. Our approach leverages a generator to produce explanations for the original input graphs while incorporating a discriminator to oversee the generation process, ensuring explanation fidelity and improving accuracy. Experimental evaluations conducted on both synthetic and real-world graph datasets demonstrate the superiority of our proposed method compared to other existing GNN explainers.

Suicide is a critical global issue that demands a comprehensive examination of factors such as mental illness, substance abuse, financial stress, and trauma. Effectively identifying individuals at risk is vital for intervention and prevention efforts. However, distinguishing suicidal ideation (SID) from non-suicidal language poses challenges. Existing research has addressed this issue, but limited attention has been given to visually interpretable and interactive systems tailored for SID. This study contributes to responsible AI by leveraging deep learning and machine learning techniques to enhance SID detection, enabling proactive interventions and support. In this paper, we introduce SIDVis, an interactive visualization system that improves performance and interpretability at the same time. The rigorous evaluation demonstrates that SIDVis not only outperforms existing methods in terms of accuracy but also provides an explanation for the responsible use of the underlying AI approach, demonstrating its potential to improve SID detection and intervention strategies.

As an extension of the Internet of things (IoT), Internet of vehicles (IoV) paradigm plays a crucial role in advancing the development of smart cities. IoV relies on vehicular communication, enabling real-time interactions between vehicles, roadside infrastructure, and pedestrians. The main goal of IoV is not only to enhance road safety services but also to support time-sensitive IoT applications. In this context, cellular vehicle-to-everything (C-V2X) communication emerges as a prominent technology for achieving IoV goals. However, C-V2X communication system faces great challenges due to spectrum scarcity and the requirement for low-latency communication in high mobility and dynamic channel conditions. To meet these challenges, cognitive radio (CR)-inspired nonorthogonal multiple access (NOMA) has emerged as a promising solution to enhance the system capacity and spectrum efficiency. In this paper, we propose a novel CR paradigm, which utilizes the spectrum holes in an embedded mode. Namely, the secondary users (SUs) are allowed to access the holes released by idle primary users (PUs) without degrading the performance of active PUs. In addition, considering the channel aging phenomenon, we perform channel prediction to reduce the performance degradation. An online learning-based scheme that enables real-time resource allocation within the embedded CR assisted NOMA framework is then designed. Simulation results demonstrate superior performance gain of the proposed scheme. When compared with the conventional NOMA, the assistance of CR brings around threefold capacity gain, and when compared with the random allocation scheme, the capacity is increased by 21%.

The recent Coronavirus Infectious Disease 2019 (COVID-19) pandemic has caused an unprecedented impact across the globe. We have also witnessed millions of people with increased mental health issues, such as depression, stress, worry, fear, disgust, sadness, and anxiety, which have become one of the major public health concerns during this severe health crisis. Depression can cause serious emotional, behavioral, and physical health problems with significant consequences, both personal and social costs included. This article studies community depression dynamics due to the COVID-19 pandemic through user-generated content on Twitter. A new approach based on multimodal features from tweets and term frequency-inverse document frequency (TF-IDF) is proposed to build depression classification models. Multimodal features capture depression cues from emotion, topic, and domain-specific perspectives. We study the problem using recently scraped tweets from Twitter users emanating from the state of New South Wales in Australia. Our novel classification model is capable of extracting depression polarities that may be affected by COVID-19 and related events during the COVID-19 period. The results found that people became more depressed after the outbreak of COVID-19. The measures implemented by the government, such as the state lockdown, also increased depression levels.

Diagram-sentence matching is a valuable academic research because it can help learners effectively understand the diagrams with the assisted by sentences. However, there are many uncommon objects, i.e. few-shot contents in diagrams and sentences. The existing methods for image-sentence matching have great limitations when applied to diagrams. Because they focus on the high-frequency objects during training and ignore the uncommon objects. In addition, the specialty leads to the semantic non-intuition of the diagram itself. In this work, we propose a cross-modal attention graph model for the few-shot diagram-sentence matching task named Fs-DSM. Specifically, it is composed of three modules. The graph initialization module regards the region-level diagram features and word-level sentence features as the nodes of Fs-DSM, and edges are represented as similarity between nodes. The information propagation module is a key point of Fs-DSM, in which the few-shot contents are recognized by an uncommon object recognition strategy, and then the nodes are updated by a neighborhood aggregation procedure with cross-modal propagation between all visual and textual nodes, while the edges are recomputed based on the new node features. The global association module integrates the features of regions and words to represent the global diagrams and sentences. By conducting comprehensive experiments in terms of few-shot and conventional image-sentence matching, we demonstrate that Fs-DSM achieves superior performances over the competitors on the AI2D [Formula: see text] diagram dataset and two public benchmark datasets with nature images.

Machine Learning (ML) plays a key role in various intelligent systems, and building an effective ML model for a data set is a difficult task involving various steps including data cleaning, feature definition and extraction, ML algorithms development, model training and evaluation as well as others. One of the most important steps in the process is to compare generated substantial amounts of ML models to find the optimal one for the deployment. It is challenging to compare such models with dynamic number of features. This paper proposes a novel visualisation approach based on a radial net to compare ML models trained with a different number of features of a given data set while revealing implicit dependent relations. In the proposed approach, ML models and features are represented by lines and arcs respectively. The dependence of ML models with dynamic number of features is encoded into the structure of visualisation, where ML models and their dependent features are directly revealed from related line connections. ML model performance information is encoded with colour and line width in the innovative visualisation. Together with the structure of visualization, feature importance can be directly discerned to help to understand ML models.

In the human-machine collaboration context, understanding the reason behind each human decision is critical for interpreting the performance of the human-machine team. Via an experimental study of a system with varied levels of accuracy, we describe how human trust interplays with system performance, human perception and decisions. It is revealed that humans are able to perceive the performance of automatic systems and themselves, and adjust their trust levels according to the accuracy of systems. The 70% system accuracy suggests to be a threshold between increasing and decreasing human trust and system usage. We have also shown that trust can be derived from a series of users' decisions rather than from a single one, and relates to the perceptions of users. A general framework depicting how trust and perception affect human decision making is proposed, which can be used as future guidelines for human-machine collaboration design.

This paper investigates End-User Development (EUD) for interactive data-analytic interfaces-building upon the ideas of making machine learning transparent. The research is carried out in a business operation environment (water pipe failure prediction in our case) motivated to integrate advanced analytics into decision-making processes of an urban Internet of Things (IoT) concept. We explore effects of revealing uncertainty and correlation on user confidence in a data-driven decision making scenario. It was found that user confidence varied significantly amongst various user groups when different machine learning models were displayed with/without supplementary information. Galvanic Skin Response (GSR) signals were analyzed and shown as reasonable indices for predicting user confidence levels. Supplementary data visualizations (of inherent uncertainty and correlation in data) contributed to explicability principles while GSR indexing added towards correctibility principles. We recommend transparent machine learning as the key to effective EUD for interactive data analytics.

As commercial pen-centric systems proliferate, they create a parallel need for analytic techniques based on dynamic writing. Within educational applications, recent empirical research has shown that signal-level features of students’ writing, such as stroke distance, pressure and duration, are adapted to conserve total energy expenditure as they consolidate expertise in a domain. The present research examined how accurately three different machine-learning algorithms could automatically classify users’ domain expertise based on signal features of their writing, without any content analysis. Compared with an unguided machine-learning classification accuracy of 71%, hybrid methods using empirical-statistical guidance correctly classified 79–92% of students by their domain expertise level. In addition to improved accuracy, the hybrid approach contributed a causal understanding of prediction success and generalization to new data. These novel findings open up opportunities to design new automated learning analytic systems and student-adaptive educational technologies for the rapidly expanding sector of commercial pen systems.

Safety-oriented visualization is one of significant approaches to gain insights from time-spatial data while neural net currently serves as a decent way to perform machine learning in data mining industry. This paper proposes a visual analytics pipeline for trajectory data enabling better understanding movements pattern of people using Neural Network as back-end and other visualization techniques as front-end for gaining information of preferences of attractions, similarities of groups, popularities of attractions and pattern of movement flow. Such understandings help to address the management issue by extracting the outstanding features to detect abnormal pattern such as detection of crime and predicting overall movements, and so on. Successfully dealing with those issues would have significant improvements of entire management of public facility such as parks and transportation.

Trust is a key factor affecting the way people rely on automated systems. On the other hand, system performance has comprehensive implications on a user's trust variations. This paper examines systems of varied levels of accuracy, in order to reveal the relationship between system performance, a user's trust and reliance on the system. In particular, it is identified that system failures have a stronger effect on trust than system successes. We also describe how patterns of trust change according to a number of consecutive system failures or successes. Importantly, we show that increasing user familiarity with the system decreases the rate of trust change, which provides new insights on the development of user trust. Finally, our analysis established a correlation between a user's reliance on a system and their trust level. Combining all these findings can have important implications in general system design and implementation, by predicting how trust builds and when it stabilizes, as well as allowing for indirectly reading a user's trust in real time based on system reliance.

Trust plays an important role in various user-facing systems and applications. It is particularly important in the context of decision support systems, where the system's output serves as one of the inputs for the users' decision making processes. In this work, we study the dynamics of explicit and implicit user trust in a simulated automated quality monitoring system, as a function of the system accuracy. We establish that users correctly perceive the accuracy of the system and adjust their trust accordingly.

BACKGROUND: Direct volume rendering is one of flexible and effective approaches to inspect large volumetric data such as medical and biological images. In conventional volume rendering, it is often time consuming to set up a meaningful illumination environment. Moreover, conventional illumination approaches usually assign same values of variables of an illumination model to different structures manually and thus neglect the important illumination variations due to structure differences. RESULTS: We introduce a novel illumination design paradigm for volume rendering on the basis of topology to automate illumination parameter definitions meaningfully. The topological features are extracted from the contour tree of an input volumetric data. The automation of illumination design is achieved based on four aspects of attenuation, distance, saliency, and contrast perception. To better distinguish structures and maximize illuminance perception differences of structures, a two-phase topology-aware illuminance perception contrast model is proposed based on the psychological concept of Just-Noticeable-Difference. CONCLUSIONS: The proposed approach allows meaningful and efficient automatic generations of illumination in volume rendering. Our results showed that our approach is more effective in depth and shape depiction, as well as providing higher perceptual differences between structures.

Prevailing social norms prohibit interrupting another person when they are speaking. In this research, simultaneous speech was investigated in groups of students as they jointly solved math problems and peer tutored one another. Analyses were based on the Math Data Corpus, which includes ground-truth performance coding and speech transcriptions. Simultaneous speech was elevated 120-143% during the most productive phase of problem solving, compared with matched intervals. It also was elevated 18-37% in students who were domain experts, compared with non-experts. Qualitative analyses revealed that experts differed from non-experts in the function of their interruptions. Analysis of these functional asymmetries produced nine key behaviors that were used to identify the dominant math expert in a group with 95-100% accuracy in three minutes. This research demonstrates that overlapped speech is a marker of group problem-solving progress and domain expertise. It provides valuable information for the emerging field of learning analytics.

This article presents a framework of adaptive, measurable decision making for Multiple Attribute Decision Making (MADM) by varying decision factors in their types, numbers, and values. Under this framework, decision making is measured using physiological sensors such as Galvanic Skin Response (GSR) and eye-tracking while users are subjected to varying decision quality and difficulty levels. Following this quantifiable decision making, users are allowed to refine several decision factors in order to make decisions of high quality and with low difficulty levels. A case study of driving route selection is used to set up an experiment to test our hypotheses. In this study, GSR features exhibit the best performance in indexing decision quality. These results can be used to guide the design of intelligent user interfaces for decision-related applications in HCI that can adapt to user behavior and decision-making performance.

Lighting design plays critical roles in depicting structural details in volume rendering. Insufficient and excessive illumination can both affect effectiveness of presenting structural details in visualization. This paper introduces topological importance into the lighting design and proposes the importance-aware lighting. In the proposed approach, the lighting in volume rendering is enhanced based on topological importance. As a result, importance of structures can be depicted from the lighting perspective. The contour tree, one of topological data structures, is used to represent topology in this paper. Topological importance such as persistence derived from the contour tree is used to modulate lighting coefficients. The experimental results demonstrate that the importance-aware lighting not only helps to depict structural details more clearly but also reveal topological importance of structures in rendering. The importance-aware lighting is more meaningful to users but not a random selection without physical meanings based on preferences.

In a wide range of biological studies, it is highly desirable to visualize and analyze three-dimensional (3D) microscopic images. In this primer, we first introduce several major methods for visualizing typical 3D images and related multi-scale, multi-time-point, multi-color data sets. Then, we discuss three key categories of image analysis tasks, namely segmentation, registration, and annotation. We demonstrate how to pipeline these visualization and analysis modules using examples of profiling the single-cell gene-expression of C. elegans and constructing a map of stereotyped neurite tracts in a fruit fly brain.

Transfer functions facilitate the volumetric data visualization by assigning optical properties to various data features and scalar values. Automation of transfer function specifications still remains a challenge in volume rendering. This paper presents an approach for automating transfer function generations by utilizing topological attributes derived from the contour tree of a volume. The contour tree acts as a visual index to volume segments, and captures associated topological attributes involved in volumetric data. A residue flow model based on Darcy's Law is employed to control distributions of opacity between branches of the contour tree. Topological attributes are also used to control color selection in a perceptual color space and create harmonic color transfer functions. The generated transfer functions can depict inclusion relationship between structures and maximize opacity and color differences between them. The proposed approach allows efficient automation of transfer function generations, and exploration on the data to be carried out based on controlling of opacity residue flow rate instead of complex low-level transfer function parameter adjustments. Experiments on various data sets demonstrate the practical use of our approach in transfer function generations.

In this paper, a new approach named focal region-based volume rendering for visualizing internal structures of volumetric data is presented. This approach presents volumetric information through integrating context information as the structure analysis of the data set with a lens-like focal region rendering to show more detailed information. This feature-based approach contains three main components: (i) A feature extraction model using 3D image processing techniques to explore the structure of objects to provide contextual information; (ii) An efficient ray-bounded volume ray casting rendering to provide the detailed information of the volume of interest in the focal region; (iii) The tools used to manipulate focal regions to make this approach more flexible. The approach provides a powerful framework for producing detailed information from volumetric data. Providing contextual information and focal region renditions at the same time has the advantages of easy to understand and comprehend volume information for the scientist. The interaction techniques provided in this approach make the focal region-based volume rendering more flexible and easy to use.