Towards Better Analysis of Deep Convolutional Neural Networks

Automatic mandibular third molar and mandibular canal relationship determination based on deep learning models for preoperative risk reduction

OBJECTIVES

This study explores the application of deep learning models for classifying the spatial relationship between mandibular third molars and the mandibular canal using cone-beam computed tomography images. Accurate classification of this relationship is essential for preoperative planning, as improper assessment can lead to complications such as inferior alveolar nerve injury during extractions.

MATERIALS AND METHODS

A dataset of 305 cone-beam computed tomography scans, categorized into three classes (not contacted, nearly contacted, and contacted), was meticulously annotated and validated by maxillofacial radiology experts to ensure reliability. Multiple state-of-the-art convolutional neural networks, including MobileNet, Xception, and DenseNet201, were trained and evaluated. Performance metrics were analysed.

RESULTS

MobileNet achieved the highest overall performance, with an accuracy of 99.44%. Xception and DenseNet201 also demonstrated strong classification capabilities, with accuracies of 98.74% and 98.73%, respectively.

CONCLUSIONS

These results highlight the potential of deep learning models to automate and improve the accuracy and consistency of mandibular third molars and the mandibular canal relationship classifications.

CLINICAL RELEVANCE

The integration of such systems into clinical workflows could enhance surgical risk assessments, streamline diagnostics, and reduce reliance on manual analysis, particularly in resource-constrained settings. This study contributes to advancing the use of artificial intelligence in dental imaging, offering a promising avenue for safer and more efficient surgical planning.

Genetic improvement of low-lignin poplars: a new strategy based on molecular recognition, chemical reactions and empirical breeding

As an important source of pollution in the papermaking process, the presence of lignin in poplar can seriously affect the quality and process of pulping. During lignin synthesis, Caffeoyl-CoA-O methyltransferase (CCoAOMT), as a specialized catalytic transferase, can effectively regulate the methylation of caffeoyl-coenzyme A (CCoA) to feruloyl-coenzyme A. Targeting CCoAOMT, this study investigated the substrate recognition mechanism and the possible reaction mechanism, the key residues of lignin binding were mutated and the lignin content was validated by deep convolutional neural-network model based on genome-wide prediction (DCNGP). The molecular mechanics results indicate that the binding of S-adenosyl methionine (SAM) and CCoA is sequential, with SAM first binding and inducing inward constriction of the CCoAOMT; then CCoA binds to the pocket, and this process closes the outer channel, preventing contamination by impurities and ensuring that the reaction proceeds. Next, the key residues in the recognition process of SAM (F69 and D91) and CCoA (I40, N170, Y188 and D218) were analyzed, and we identified that K146 as a base catalyst is important for inducing the methylation reaction. Immediately after that, the possible methylation reaction mechanism was deduced by the combination of Restrained Electrostatic Potential (RESP) and Independent Gradient Model (IGM) analysis, focusing on the catalytic center electron cloud density and RESP charge distribution. Finally, the DCNGP results verified that the designed mutant groups were all able to effectively reduce the lignin content and increase the S-lignin content/ G-lignin content ratio, which was beneficial for the subsequent lignin removal. Multifaceted consideration of factors that reduce lignin content and combined deep learning to screen for favorable mutations in target traits provides new ideas for targeted breeding of low-lignin poplars.

Evaluating and Extending Speedup Techniques for Optimal Crossing Minimization in Layered Graph Drawings

A layered graph is an important category of graph in which every node is assigned to a layer, and layers are drawn as parallel or radial lines. They are commonly used to display temporal data or hierarchical graphs. Previous research has demonstrated that minimizing edge crossings is the most important criterion to consider when looking to improve the readability of such graphs. While heuristic approaches exist for crossing minimization, we are interested in optimal approaches to the problem that prioritize human readability over computational scalability. We aim to improve the usefulness and applicability of such optimal methods by understanding and improving their scalability to larger graphs. This paper categorizes and evaluates the state-of-the-art linear programming formulations for exact crossing minimization and describes nine new and existing techniques that could plausibly accelerate the optimization algorithm. Through a computational evaluation, we explore each technique's effect on calculation time and how the techniques assist or inhibit one another, allowing researchers and practitioners to adapt them to the characteristics of their graphs. Our best-performing techniques yielded a median improvement of 2.5-17 × depending on the solver used, giving us the capability to create optimal layouts faster and for larger graphs. We provide an open-source implementation of our methodology in Python, where users can pick which combination of techniques to enable according to their use case. A free copy of this paper and all supplemental materials, datasets used, and source code are available at https://osf.io/5vq79.

Dynamic Color Assignment for Hierarchical Data

Assigning discriminable and harmonic colors to samples according to their class labels and spatial distribution can generate attractive visualizations and facilitate data exploration. However, as the number of classes increases, it is challenging to generate a high-quality color assignment result that accommodates all classes simultaneously. A practical solution is to organize classes into a hierarchy and then dynamically assign colors during exploration. However, existing color assignment methods fall short in generating high-quality color assignment results and dynamically aligning them with hierarchical structures. To address this issue, we develop a dynamic color assignment method for hierarchical data, which is formulated as a multi-objective optimization problem. This method simultaneously considers color discriminability, color harmony, and spatial distribution at each hierarchical level. By using the colors of parent classes to guide the color assignment of their child classes, our method further promotes both consistency and clarity across hierarchical levels. We demonstrate the effectiveness of our method in generating dynamic color assignment results with quantitative experiments and a user study.

AdversaFlow: Visual Red Teaming for Large Language Models with Multi-Level Adversarial Flow

Large Language Models (LLMs) are powerful but also raise significant security concerns, particularly regarding the harm they can cause, such as generating fake news that manipulates public opinion on social media and providing responses to unethical activities. Traditional red teaming approaches for identifying AI vulnerabilities rely on manual prompt construction and expertise. This paper introduces AdversaFlow, a novel visual analytics system designed to enhance LLM security against adversarial attacks through human-AI collaboration. AdversaFlow involves adversarial training between a target model and a red model, featuring unique multi-level adversarial flow and fluctuation path visualizations. These features provide insights into adversarial dynamics and LLM robustness, enabling experts to identify and mitigate vulnerabilities effectively. We present quantitative evaluations and case studies validating our system's utility and offering insights for future AI security solutions. Our method can enhance LLM security, supporting downstream scenarios like social media regulation by enabling more effective detection, monitoring, and mitigation of harmful content and behaviors.

An In-Situ Visual Analytics Framework for Deep Neural Networks

The past decade has witnessed the superior power of deep neural networks (DNNs) in applications across various domains. However, training a high-quality DNN remains a non-trivial task due to its massive number of parameters. Visualization has shown great potential in addressing this situation, as evidenced by numerous recent visualization works that aid in DNN training and interpretation. These works commonly employ a strategy of logging training-related data and conducting post-hoc analysis. Based on the results of offline analysis, the model can be further trained or fine-tuned. This strategy, however, does not cope with the increasing complexity of DNNs, because (1) the time-series data collected over the training are usually too large to be stored entirely; (2) the huge I/O overhead significantly impacts the training efficiency; (3) post-hoc analysis does not allow rapid human-interventions (e.g., stop training with improper hyper-parameter settings to save computational resources). To address these challenges, we propose an in-situ visualization and analysis framework for the training of DNNs. Specifically, we employ feature extraction algorithms to reduce the size of training-related data in-situ and use the reduced data for real-time visual analytics. The states of model training are disclosed to model designers in real-time, enabling human interventions on demand to steer the training. Through concrete case studies, we demonstrate how our in-situ framework helps deep learning experts optimize DNNs and improve their analysis efficiency.

Towards Better Modeling With Missing Data: A Contrastive Learning-Based Visual Analytics Perspective

Missing data can pose a challenge for machine learning (ML) modeling. To address this, current approaches are categorized into feature imputation and label prediction and are primarily focused on handling missing data to enhance ML performance. These approaches rely on the observed data to estimate the missing values and therefore encounter three main shortcomings in imputation, including the need for different imputation methods for various missing data mechanisms, heavy dependence on the assumption of data distribution, and potential introduction of bias. This study proposes a Contrastive Learning (CL) framework to model observed data with missing values, where the ML model learns the similarity between an incomplete sample and its complete counterpart and the dissimilarity between other samples. Our proposed approach demonstrates the advantages of CL without requiring any imputation. To enhance interpretability, we introduce CIVis, a visual analytics system that incorporates interpretable techniques to visualize the learning process and diagnose the model status. Users can leverage their domain knowledge through interactive sampling to identify negative and positive pairs in CL. The output of CIVis is an optimized model that takes specified features and predicts downstream tasks. We provide two usage scenarios in regression and classification tasks and conduct quantitative experiments, expert interviews, and a qualitative user study to demonstrate the effectiveness of our approach. In short, this study offers a valuable contribution to addressing the challenges associated with ML modeling in the presence of missing data by providing a practical solution that achieves high predictive accuracy and model interpretability.

Assessment of artificial intelligence to detect gasoline in fire debris using HS-SPME-GC/MS and transfer learning

Due to the complex nature of the chemical compositions of ignitable liquids (IL) and the interferences from fire debris matrices, interpreting chromatographic data poses challenges to analysts. In this work, artificial intelligence (AI) was developed by transfer learning in a convolutional neural network (CNN), GoogLeNet. The image classification AI was fine-tuned to create intelligent classification systems to discriminate samples containing gasoline residues from burned substrates. All ground truth samples were analyzed by headspace solid-phase microextraction (HS-SPME) coupled with a gas chromatograph and mass spectrometer (GC/MS). The HS-SPME-GC/MS data were transformed into three types of image presentations, that is, heatmaps, extracted ion heatmaps, and total ion chromatograms. The abundance and mass-to-charge ratios of each scan were converted into image patterns that are characteristic of the chemical profiles of gasoline. The transfer learning data were labeled as "gasoline present" and "gasoline absent" classes. The assessment results demonstrated that all AI models achieved 100 ± 0% accuracy in identifying neat gasoline. When the models were assessed using the spiked samples, the AI model developed using the extracted ion heatmap obtained the highest accuracy rate (95.9 ± 0.4%), which was greater than those obtained by other machine learning models, ranging from 17.3 ± 0.7% to 78.7 ± 0.7%. The proposed work demonstrated that the heatmaps created from GC/MS data can represent chemical features from the samples. Additionally, the pretrained CNN models are readily available in the transfer learning workflow to develop AI for GC/MS data interpretation in fire debris analysis.

Enhancing genome-wide populus trait prediction through deep convolutional neural networks

As a promising model, genome-based plant breeding has greatly promoted the improvement of agronomic traits. Traditional methods typically adopt linear regression models with clear assumptions, neither obtaining the linkage between phenotype and genotype nor providing good ideas for modification. Nonlinear models are well characterized in capturing complex nonadditive effects, filling this gap under traditional methods. Taking populus as the research object, this paper constructs a deep learning method, DCNGP, which can effectively predict the traits including 65 phenotypes. The method was trained on three datasets, and compared with other four classic models-Bayesian ridge regression (BRR), Elastic Net, support vector regression, and dualCNN. The results show that DCNGP has five typical advantages in performance: strong prediction ability on multiple experimental datasets; the incorporation of batch normalization layers and Early-Stopping technology enhancing the generalization capabilities and prediction stability on test data; learning potent features from the data and thus circumventing the tedious steps of manual production; the introduction of a Gaussian Noise layer enhancing predictive capabilities in the case of inherent uncertainties or perturbations; fewer hyperparameters aiding to reduce tuning time across datasets and improve auto-search efficiency. In this way, DCNGP shows powerful predictive ability from genotype to phenotype, which provide an important theoretical reference for building more robust populus breeding programs.

MiMICRI: Towards Domain-centered Counterfactual Explanations of Cardiovascular Image Classification Models

The recent prevalence of publicly accessible, large medical imaging datasets has led to a proliferation of artificial intelligence (AI) models for cardiovascular image classification and analysis. At the same time, the potentially significant impacts of these models have motivated the development of a range of explainable AI (XAI) methods that aim to explain model predictions given certain image inputs. However, many of these methods are not developed or evaluated with domain experts, and explanations are not contextualized in terms of medical expertise or domain knowledge. In this paper, we propose a novel framework and python library, MiMICRI, that provides domain-centered counterfactual explanations of cardiovascular image classification models. MiMICRI helps users interactively select and replace segments of medical images that correspond to morphological structures. From the counterfactuals generated, users can then assess the influence of each segment on model predictions, and validate the model against known medical facts. We evaluate this library with two medical experts. Our evaluation demonstrates that a domain-centered XAI approach can enhance the interpretability of model explanations, and help experts reason about models in terms of relevant domain knowledge. However, concerns were also surfaced about the clinical plausibility of the counterfactuals generated. We conclude with a discussion on the generalizability and trustworthiness of the MiMICRI framework, as well as the implications of our findings on the development of domain-centered XAI methods for model interpretability in healthcare contexts.

Visualizing and Comparing Machine Learning Predictions to Improve Human-AI Teaming on the Example of Cell Lineage

We visualize the predictions of multiple machine learning models to help biologists as they interactively make decisions about cell lineage-the development of a (plant) embryo from a single ovum cell. Based on a confocal microscopy dataset, traditionally biologists manually constructed the cell lineage, starting from this observation and reasoning backward in time to establish their inheritance. To speed up this tedious process, we make use of machine learning (ML) models trained on a database of manually established cell lineages to assist the biologist in cell assignment. Most biologists, however, are not familiar with ML, nor is it clear to them which model best predicts the embryo's development. We thus have developed a visualization system that is designed to support biologists in exploring and comparing ML models, checking the model predictions, detecting possible ML model mistakes, and deciding on the most likely embryo development. To evaluate our proposed system, we deployed our interface with six biologists in an observational study. Our results show that the visual representations of machine learning are easily understandable, and our tool, LineageD+, could potentially increase biologists' working efficiency and enhance the understanding of embryos.

Explainable deep learning in plant phenotyping

The increasing human population and variable weather conditions, due to climate change, pose a threat to the world's food security. To improve global food security, we need to provide breeders with tools to develop crop cultivars that are more resilient to extreme weather conditions and provide growers with tools to more effectively manage biotic and abiotic stresses in their crops. Plant phenotyping, the measurement of a plant's structural and functional characteristics, has the potential to inform, improve and accelerate both breeders' selections and growers' management decisions. To improve the speed, reliability and scale of plant phenotyping procedures, many researchers have adopted deep learning methods to estimate phenotypic information from images of plants and crops. Despite the successful results of these image-based phenotyping studies, the representations learned by deep learning models remain difficult to interpret, understand, and explain. For this reason, deep learning models are still considered to be black boxes. Explainable AI (XAI) is a promising approach for opening the deep learning model's black box and providing plant scientists with image-based phenotypic information that is interpretable and trustworthy. Although various fields of study have adopted XAI to advance their understanding of deep learning models, it has yet to be well-studied in the context of plant phenotyping research. In this review article, we reviewed existing XAI studies in plant shoot phenotyping, as well as related domains, to help plant researchers understand the benefits of XAI and make it easier for them to integrate XAI into their future studies. An elucidation of the representations within a deep learning model can help researchers explain the model's decisions, relate the features detected by the model to the underlying plant physiology, and enhance the trustworthiness of image-based phenotypic information used in food production systems.

Dense D2C-Net: dense connection network for display-to-camera communications

In this paper, we introduce a novel Dense D2C-Net, an unobtrusive display-to-camera (D2C) communication scheme that embeds and extracts additional data via visual content through a deep convolutional neural network (DCNN). The encoding process of Dense D2C-Net establishes connections among all layers of the cover image, and fosters feature reuse to maintain the visual quality of the image. The Y channel is employed to embed binary data owing to its resilience against distortion from image compression and its lower sensitivity to color transformations. The encoder structure integrates hybrid layers that combine feature maps from the cover image and input binary data to efficiently hide the embedded data, while the addition of multiple noise layers effectively mitigates distortions caused by the optical wireless channel on the transmitted data. At the decoder, a series of 2D convolutional layers is used for extracting output binary data from the captured image. We conducted experiments in a real-world setting using a smartphone camera and a digital display, demonstrating superior performance from the proposed scheme compared to conventional DCNN-based D2C schemes across varying parameters such as transmission distance, capture angle, display brightness, and camera resolution.

Prediction and Visual Analysis of Food Safety Risk Based on TabNet-GRA

Food safety risk prediction is crucial for timely hazard detection and effective control. This study proposes a novel risk prediction method for food safety called TabNet-GRA, which combines a specialized deep learning architecture for tabular data (TabNet) with a grey relational analysis (GRA) to predict food safety risk. Initially, this study employed a GRA to derive comprehensive risk values from fused detection data. Subsequently, a food safety risk prediction model was constructed based on TabNet, and training was performed using the detection data as inputs and the comprehensive risk values calculated via the GRA as the expected outputs. Comparative experiments with six typical models demonstrated the superior fitting ability of the TabNet-based prediction model. Moreover, a food safety risk prediction and visualization system (FSRvis system) was designed and implemented based on TabNet-GRA to facilitate risk prediction and visual analysis. A case study in which our method was applied to a dataset of cooked meat products from a Chinese province further validated the effectiveness of the TabNet-GRA method and the FSRvis system. The method can be applied to targeted risk assessment, hazard identification, and early warning systems to strengthen decision making and safeguard public health by proactively addressing food safety risks.

Multi-Object Tracking in Heterogeneous environments (MOTHe) for animal video recordings

Aerial imagery and video recordings of animals are used for many areas of research such as animal behaviour, behavioural neuroscience and field biology. Many automated methods are being developed to extract data from such high-resolution videos. Most of the available tools are developed for videos taken under idealised laboratory conditions. Therefore, the task of animal detection and tracking for videos taken in natural settings remains challenging due to heterogeneous environments. Methods that are useful for field conditions are often difficult to implement and thus remain inaccessible to empirical researchers. To address this gap, we present an open-source package called Multi-Object Tracking in Heterogeneous environments (MOTHe), a Python-based application that uses a basic convolutional neural network for object detection. MOTHe offers a graphical interface to automate the various steps related to animal tracking such as training data generation, animal detection in complex backgrounds and visually tracking animals in the videos. Users can also generate training data and train a new model which can be used for object detection tasks for a completely new dataset. MOTHe doesn't require any sophisticated infrastructure and can be run on basic desktop computing units. We demonstrate MOTHe on six video clips in varying background conditions. These videos are from two species in their natural habitat-wasp colonies on their nests (up to 12 individuals per colony) and antelope herds in four different habitats (up to 156 individuals in a herd). Using MOTHe, we are able to detect and track individuals in all these videos. MOTHe is available as an open-source GitHub repository with a detailed user guide and demonstrations at: https://github.com/tee-lab/MOTHe-GUI.

How Does Attention Work in Vision Transformers? A Visual Analytics Attempt

Vision transformer (ViT) expands the success of transformer models from sequential data to images. The model decomposes an image into many smaller patches and arranges them into a sequence. Multi-head self-attentions are then applied to the sequence to learn the attention between patches. Despite many successful interpretations of transformers on sequential data, little effort has been devoted to the interpretation of ViTs, and many questions remain unanswered. For example, among the numerous attention heads, which one is more important? How strong are individual patches attending to their spatial neighbors in different heads? What attention patterns have individual heads learned? In this work, we answer these questions through a visual analytics approach. Specifically, we first identify what heads are more important in ViTs by introducing multiple pruning-based metrics. Then, we profile the spatial distribution of attention strengths between patches inside individual heads, as well as the trend of attention strengths across attention layers. Third, using an autoencoder-based learning solution, we summarize all possible attention patterns that individual heads could learn. Examining the attention strengths and patterns of the important heads, we answer why they are important. Through concrete case studies with experienced deep learning experts on multiple ViTs, we validate the effectiveness of our solution that deepens the understanding of ViTs from head importance, head attention strength, and head attention pattern.

The Flow of Trust: A Visualization Framework to Externalize, Explore, and Explain Trust in ML Applications

We present a conceptual framework for the development of visual interactive techniques to formalize and externalize trust in machine learning (ML) workflows. Currently, trust in ML applications is an implicit process that takes place in the user's mind. As such, there is no method of feedback or communication of trust that can be acted upon. Our framework will be instrumental in developing interactive visualization approaches that will help users to efficiently and effectively build and communicate trust in ways that fit each of the ML process stages. We formulate several research questions and directions that include: 1) a typology/taxonomy of trust objects, trust issues, and possible reasons for (mis)trust; 2) formalisms to represent trust in machine-readable form; 3) means by which users can express their state of trust by interacting with a computer system (e.g., text, drawing, marking); 4) ways in which a system can facilitate users' expression and communication of the state of trust; and 5) creation of visual interactive techniques for representation and exploration of trust over all stages of an ML pipeline.

SIGN: Statistical Inference Graphs Based on Probabilistic Network Activity Interpretation

Convolutional neural networks (CNNs) have achieved superior accuracy in many visual-related tasks. However, the inference process through a CNN's intermediate layers is opaque, making it difficult to interpret such networks or develop trust in their operation. In this article, we introduce SIGN method for modeling the network's hidden layer activity using probabilistic models. The activity patterns in layers of interest are modeled as Gaussian mixture models, and transition probabilities between clusters in consecutive modeled layers are estimated to identify paths of inference. For fully connected networks, the entire layer activity is clustered, and the resulting model is a hidden Markov model. For convolutional layers, spatial columns of activity are clustered, and a maximum likelihood model is developed for mining an explanatory inference graph. The graph describes the hierarchy of activity clusters most relevant for network prediction. We show that such inference graphs are useful for understanding the general inference process of a class, as well as explaining the (correct or incorrect) decisions the network makes about specific images. In addition, SIGN provide interesting observations regarding hidden layer activity in general, including the concentration of memorization in a single middle layer in fully connected networks, and a highly local nature of column activities in the top CNN layers.

A Visual Analytics System for Improving Attention-based Traffic Forecasting Models

With deep learning (DL) outperforming conventional methods for different tasks, much effort has been devoted to utilizing DL in various domains. Researchers and developers in the traffic domain have also designed and improved DL models for forecasting tasks such as estimation of traffic speed and time of arrival. However, there exist many challenges in analyzing DL models due to the black-box property of DL models and complexity of traffic data (i.e., spatio-temporal dependencies). Collaborating with domain experts, we design a visual analytics system, AttnAnalyzer, that enables users to explore how DL models make predictions by allowing effective spatio-temporal dependency analysis. The system incorporates dynamic time warping (DTW) and Granger causality tests for computational spatio-temporal dependency analysis while providing map, table, line chart, and pixel views to assist user to perform dependency and model behavior analysis. For the evaluation, we present three case studies showing how AttnAnalyzer can effectively explore model behaviors and improve model performance in two different road networks. We also provide domain expert feedback.

Visual Analysis of Neural Architecture Spaces for Summarizing Design Principles

Recent advances in artificial intelligence largely benefit from better neural network architectures. These architectures are a product of a costly process of trial-and-error. To ease this process, we develop ArchExplorer, a visual analysis method for understanding a neural architecture space and summarizing design principles. The key idea behind our method is to make the architecture space explainable by exploiting structural distances between architectures. We formulate the pairwise distance calculation as solving an all-pairs shortest path problem. To improve efficiency, we decompose this problem into a set of single-source shortest path problems. The time complexity is reduced from O(kn²N) to O(knN). Architectures are hierarchically clustered according to the distances between them. A circle-packing-based architecture visualization has been developed to convey both the global relationships between clusters and local neighborhoods of the architectures in each cluster. Two case studies and a post-analysis are presented to demonstrate the effectiveness of ArchExplorer in summarizing design principles and selecting better-performing architectures.

In Defence of Visual Analytics Systems: Replies to Critics

The last decade has witnessed many visual analytics (VA) systems that make successful applications to wide-ranging domains like urban analytics and explainable AI. However, their research rigor and contributions have been extensively challenged within the visualization community. We come in defence of VA systems by contributing two interview studies for gathering critics and responses to those criticisms. First, we interview 24 researchers to collect criticisms the review comments on their VA work. Through an iterative coding and refinement process, the interview feedback is summarized into a list of 36 common criticisms. Second, we interview 17 researchers to validate our list and collect their responses, thereby discussing implications for defending and improving the scientific values and rigor of VA systems. We highlight that the presented knowledge is deep, extensive, but also imperfect, provocative, and controversial, and thus recommend reading with an inclusive and critical eye. We hope our work can provide thoughts and foundations for conducting VA research and spark discussions to promote the research field forward more rigorously and vibrantly.

Application of Machine Learning in Intelligent Medical Image Diagnosis and Construction of Intelligent Service Process

The introduction of digital technology in the healthcare industry is marked by ongoing difficulties with implementation and use. Slow progress has been made in unifying different healthcare systems, and much of the globe still lacks a fully integrated healthcare system. As a result, it is critical and advantageous for healthcare providers to comprehend the fundamental ideas of AI in order to design and deliver their own AI-powered technology. AI is commonly defined as the capacity of machines to mimic human cognitive functions. It can tackle jobs with equivalent or superior performance to humans by combining computer science, algorithms, machine learning, and data science. The healthcare system is a dynamic and evolving environment, and medical experts are constantly confronted with new issues, shifting duties, and frequent interruptions. Because of this variation, illness diagnosis frequently becomes a secondary concern for healthcare professionals. Furthermore, clinical interpretation of medical information is a cognitively demanding endeavor. This applies not just to seasoned experts, but also to individuals with varying or limited skills, such as young assistant doctors. In this paper, we proposed the comparative analysis of various state-of-the-art methods of deep learning for medical imaging diagnosis and evaluated various important characteristics. The methodology is to evaluate various important factors such as interpretability, visualization, semantic data, and quantification of logical relationships in medical data. Furthermore, the glaucoma diagnosis system is discussed in detail via qualitative and quantitative approaches. Finally, the applications and future prospects were also discussed.

Visual Analytics for RNN-Based Deep Reinforcement Learning

Deep reinforcement learning (DRL) targets to train an autonomous agent to interact with a pre-defined environment and strives to achieve specific goals through deep neural networks (DNN). Recurrent neural network (RNN) based DRL has demonstrated superior performance, as RNNs can effectively capture the temporal evolution of the environment and respond with proper agent actions. However, apart from the outstanding performance, little is known about how RNNs understand the environment internally and what has been memorized over time. Revealing these details is extremely important for deep learning experts to understand and improve DRLs, which in contrast, is also challenging due to the complicated data transformations inside these models. In this article, we propose Deep Reinforcement Learning Interactive Visual Explorer (DRLIVE), a visual analytics system to effectively explore, interpret, and diagnose RNN-based DRLs. Having focused on DRL agents trained for different Atari games, DRLIVE accomplishes three tasks: game episode exploration, RNN hidden/cell state examination, and interactive model perturbation. Using the system, one can flexibly explore a DRL agent through interactive visualizations, discover interpretable RNN cells by prioritizing RNN hidden/cell states with a set of metrics, and further diagnose the DRL model by interactively perturbing its inputs. Through concrete studies with multiple deep learning experts, we validated the efficacy of DRLIVE.

Inspecting the Running Process of Horizontal Federated Learning via Visual Analytics

As a decentralized training approach, horizontal federated learning (HFL) enables distributed clients to collaboratively learn a machine learning model while keeping personal/private information on local devices. Despite the enhanced performance and efficiency of HFL over local training, clues for inspecting the behaviors of the participating clients and the federated model are usually lacking due to the privacy-preserving nature of HFL. Consequently, the users can only conduct a shallow-level analysis of potential abnormal behaviors and have limited means to assess the contributions of individual clients and implement the necessary intervention. Visualization techniques have been introduced to facilitate the HFL process inspection, usually by providing model metrics and evaluation results as a dashboard representation. Although the existing visualization methods allow a simple examination of the HFL model performance, they cannot support the intensive exploration of the HFL process. In this article, strictly following the HFL privacy-preserving protocol, we design an exploratory visual analytics system for the HFL process termed HFLens, which supports comparative visual interpretation at the overview, communication round, and client instance levels. Specifically, the proposed system facilitates the investigation of the overall process involving all clients, the correlation analysis of clients' information in one or different communication round(s), the identification of potential anomalies, and the contribution assessment of each HFL client. Two case studies confirm the efficacy of our system. Experts' feedback suggests that our approach indeed helps in understanding and diagnosing the HFL process better.

Towards Systematic Design Considerations for Visualizing Cross-View Data Relationships

Due to the scale of data and the complexity of analysis tasks, insight discovery often requires coordinating multiple visualizations (views), with each view displaying different parts of data or the same data from different perspectives. For example, to analyze car sales records, a marketing analyst uses a line chart to visualize the trend of car sales, a scatterplot to inspect the price and horsepower of different cars, and a matrix to compare the transaction amounts in types of deals. To explore related information across multiple views, current visual analysis tools heavily rely on brushing and linking techniques, which may require a significant amount of user effort (e.g., many trial-and-error attempts). There may be other efficient and effective ways of displaying cross-view data relationships to support data analysis with multiple views, but currently there are no guidelines to address this design challenge. In this article, we present systematic design considerations for visualizing cross-view data relationships, which leverages descriptive aspects of relationships and usable visual context of multi-view visualizations. We discuss pros and cons of different designs for showing cross-view data relationships, and provide a set of recommendations for helping practitioners make design decisions.

Understanding the role of eye movement consistency in face recognition and autism through integrating deep neural networks and hidden Markov models

Greater eyes-focused eye movement pattern during face recognition is associated with better performance in adults but not in children. We test the hypothesis that higher eye movement consistency across trials, instead of a greater eyes-focused pattern, predicts better performance in children since it reflects capacity in developing visual routines. We first simulated visual routine development through combining deep neural network and hidden Markov model that jointly learn perceptual representations and eye movement strategies for face recognition. The model accounted for the advantage of eyes-focused pattern in adults, and predicted that in children (partially trained models) consistency but not pattern of eye movements predicted recognition performance. This result was then verified with data from typically developing children. In addition, lower eye movement consistency in children was associated with autism diagnosis, particularly autistic traits in social skills. Thus, children's face recognition involves visual routine development through social exposure, indexed by eye movement consistency.

Intelligent identification of film on cotton based on hyperspectral imaging and convolutional neural network

The identification of the film on cotton is of great significance for the improvement of cotton quality. Most of the existing technologies are dedicated to removing colored foreign fibers from cotton using photoelectric sorting methods. However, the current technologies are difficult to identify colorless transparent film, which becomes an obstacle for the harvest of high-quality cotton. In this paper, an intelligent identification method is proposed to identify the colorless and transparent film on cotton, based on short-wave near-infrared hyperspectral imaging and convolutional neural network (CNN). The algorithm includes black-and-white correction of hyperspectral images, hyperspectral data dimensionality reduction, CNN model training and testing. The key technology is that the features of the hyperspectral image data are degraded by the principal component analysis (PCA) to reduce the amount of computing time. The main innovation is that the colorless and transparent film on cotton can be accurately identified through a CNN with the performance of automatic feature extraction. The experimental results show that the proposed method can greatly improve the identification precision, compared with the traditional methods. After the simulation experiment, the method proposed in this paper has a recognition rate of 98.5% for film. After field testing, the selection rate of film is as high as 96.5%, which meets the actual production needs.

A walk in the black-box: 3D visualization of large neural networks in virtual reality

Within the last decade Deep Learning has become a tool for solving challenging problems like image recognition. Still, Convolutional Neural Networks (CNNs) are considered black-boxes, which are difficult to understand by humans. Hence, there is an urge to visualize CNN architectures, their internal processes and what they actually learn. Previously, virtual realityhas been successfully applied to display small CNNs in immersive 3D environments. In this work, we address the problem how to feasibly render large-scale CNNs, thereby enabling the visualization of popular architectures with ten thousands of feature maps and branches in the computational graph in 3D. Our software ”DeepVisionVR” enables the user to freely walk through the layered network, pick up and place images, move/scale layers for better readability, perform feature visualization and export the results. We also provide a novel Pytorch module to dynamically link PyTorch with Unity, which gives developers and researchers a convenient interface to visualize their own architectures. The visualization is directly created from the PyTorch class that defines the Pytorch model used for training and testing. This approach allows full access to the network’s internals and direct control over what exactly is visualized. In a use-case study, we apply the module to analyze models with different generalization abilities in order to understand how networks memorize images. We train two recent architectures, CovidResNet and CovidDenseNet on the Caltech101 and the SARS-CoV-2 datasets and find that bad generalization is driven by high-frequency features and the susceptibility to specific pixel arrangements, leading to implications for the practical application of CNNs. The code is available on Github https://github.com/Criscraft/DeepVisionVR.

E-Commerce Brand Ranking Algorithm Based on User Evaluation and Sentiment Analysis

Objective

Consumers often need to compare the same type of products from different merchants to determine their purchasing needs. Fully mining the product information on the website and applying it to e-commerce websites or product introduction websites can not only allow consumers to buy products that are more in line with their wishes, but also help merchants understand user needs and the advantages of each product. How to quantify the emotional tendency of evaluation information and how to recommend satisfactory products to consumers is the research purpose of this paper.

Method

According to the analysis of the research object, this paper uses the Python crawler library to efficiently crawl the data required for research. By writing a custom program for crawling, the resulting data is more in line with the actual situation. This paper uses the BeautifulSoup library in Python web crawler technology for data acquisition. Then, in order to ensure high-quality data sets, the acquired data needs to be cleaned and deduplicated. Finally, preprocessing such as sentence segmentation, word segmentation, and semantic analysis is performed on the cleaned data, and the data format required by the subsequent model is output. For weightless network, the concept of node similarity is proposed, which is used to measure the degree of mutual influence between nodes. Combined with the LeaderRank algorithm, and fully considering the differences between nodes in the interaction, the SRank algorithm is proposed. Different from the classical node importance ranking method, the SRank algorithm fully considers the local and global characteristics of nodes, which is more in line with the actual network.

Results/Discussion

This paper calculates the sentiment polarity of users’ comments, obtains the final user influence ranking, and identifies opinion leaders. The final ranking results were compared and analyzed with the traditional PageRank algorithm and SRank ranking algorithm, and it was found that the opinion leaders identified by the opinion leader identification model integrating user activity and comment sentiment were more reasonable and objective. The algorithm in this paper improves the efficiency of operation to a certain extent, and at the same time solves the problem that sentiment analysis cannot be effectively used in social network analysis, and increases the accuracy of e-commerce brand ranking.

VAC-CNN: A Visual Analytics System for Comparative Studies of Deep Convolutional Neural Networks

The rapid development of Convolutional Neural Networks (CNNs) in recent years has triggered significant breakthroughs in many machine learning (ML) applications. The ability to understand and compare various CNN models available is thus essential. The conventional approach with visualizing each model's quantitative features, such as classification accuracy and computational complexity, is not sufficient for a deeper understanding and comparison of the behaviors of different models. Moreover, most of the existing tools for assessing CNN behaviors only support comparison between two models and lack the flexibility of customizing the analysis tasks according to user needs. This paper presents a visual analytics system, VAC-CNN (Visual Analytics for Comparing CNNs), that supports the in-depth inspection of a single CNN model as well as comparative studies of two or more models. The ability to compare a larger number of (e.g., tens of) models especially distinguishes our system from previous ones. With a carefully designed model visualization and explaining support, VAC-CNN facilitates a highly interactive workflow that promptly presents both quantitative and qualitative information at each analysis stage. We demonstrate VAC-CNN's effectiveness for assisting novice ML practitioners in evaluating and comparing multiple CNN models through two use cases and one preliminary evaluation study using the image classification tasks on the ImageNet dataset.

Human-centered explainability for life sciences, healthcare, and medical informatics

Rapid advances in artificial intelligence (AI) and availability of biological, medical, and healthcare data have enabled the development of a wide variety of models. Significant success has been achieved in a wide range of fields, such as genomics, protein folding, disease diagnosis, imaging, and clinical tasks. Although widely used, the inherent opacity of deep AI models has brought criticism from the research field and little adoption in clinical practice. Concurrently, there has been a significant amount of research focused on making such methods more interpretable, reviewed here, but inherent critiques of such explainability in AI (XAI), its requirements, and concerns with fairness/robustness have hampered their real-world adoption. We here discuss how user-driven XAI can be made more useful for different healthcare stakeholders through the definition of three key personas-data scientists, clinical researchers, and clinicians-and present an overview of how different XAI approaches can address their needs. For illustration, we also walk through several research and clinical examples that take advantage of XAI open-source tools, including those that help enhance the explanation of the results through visualization. This perspective thus aims to provide a guidance tool for developing explainability solutions for healthcare by empowering both subject matter experts, providing them with a survey of available tools, and explainability developers, by providing examples of how such methods can influence in practice adoption of solutions.

Towards Better Caption Supervision for Object Detection

As training high-performance object detectors requires expensive bounding box annotations, recent methods resort to free-available image captions. However, detectors trained on caption supervision perform poorly because captions are usually noisy and cannot provide precise location information. To tackle this issue, we present a visual analysis method, which tightly integrates caption supervision with object detection to mutually enhance each other. In particular, object labels are first extracted from captions, which are utilized to train the detectors. Then, the objects detected from images are fed into caption supervision for further improvement. To effectively loop users into the object detection process, a node-link-based set visualization supported by a multi-type relational co-clustering algorithm is developed to explain the relationships between the extracted labels and the images with detected objects. The co-clustering algorithm clusters labels and images simultaneously by utilizing both their representations and their relationships. Quantitative evaluations and a case study are conducted to demonstrate the efficiency and effectiveness of the developed method in improving the performance of object detectors.

Application of an Improved Clustering Algorithm of Neural Networks in Performance Appraisal Systems

With the development of economic globalization, human resource competition has long become the key core of enterprise development and peer competition. Reasonably Formulating an enterprise’s employee performance appraisal management system and conducting standardized, fair, and just appraisal management are the basic requirements for the survival and development of an enterprise. This paper studies the application of an improved clustering algorithm based on neural network in an employee performance appraisal management system and explores its application value in the employee performance appraisal management system by using the improved ART2 clustering method that draws on leakage competition and Hebb rules. The experimental results of this paper show that the satisfaction of this system in the four aspects of integrated data management, system stability and convenience, and transparency in performance appraisal are all above 66%. This shows that this system has superior performance and good reference value.

MPTP-Treated Zebrafish Recapitulate 'Late-Stage' Parkinson's-like Cognitive Decline

The zebrafish is a promising model species in biomedical research, including neurotoxicology and neuroactive drug screening. 1-Methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP) evokes degeneration of dopaminergic neurons and is commonly used to model Parkinson's disease (PD) in laboratory animals, including zebrafish. However, cognitive phenotypes in MPTP-evoked experimental PD models remain poorly understood. Here, we established an LD50 (292 mg/kg) for intraperitoneal MPTP administration in adult zebrafish, and report impaired spatial working memory (poorer spontaneous alternation in the Y-maze) in a PD model utilizing fish treated with 200 µg of this agent. In addition to conventional behavioral analyses, we also employed artificial intelligence (AI)-based approaches to independently and without bias characterize MPTP effects on zebrafish behavior during the Y-maze test. These analyses yielded a distinct cluster for 200-μg MPTP (vs. other) groups, suggesting that high-dose MPTP produced distinct, computationally detectable patterns of zebrafish swimming. Collectively, these findings support MPTP treatment in adult zebrafish as a late-stage experimental PD model with overt cognitive phenotypes.

Artificial intelligence-driven phenotyping of zebrafish psychoactive drug responses

Zebrafish (Danio rerio) are rapidly emerging in biomedicine as promising tools for disease modelling and drug discovery. The use of zebrafish for neuroscience research is also growing rapidly, necessitating novel reliable and unbiased methods of neurophenotypic data collection and analyses. Here, we applied the artificial intelligence (AI) neural network-based algorithms to a large dataset of adult zebrafish locomotor tracks collected previously in a series of in vivo experiments with multiple established psychotropic drugs. We first trained AI to recognize various drugs from a wide range of psychotropic agents tested, and then confirmed prediction accuracy of trained AI by comparing several agents with known similar behavioral and pharmacological profiles. Presenting a framework for innovative neurophenotyping, this proof-of-concept study aims to improve AI-driven movement pattern classification in zebrafish, thereby fostering drug discovery and development utilizing this key model organism.

Revisiting Dimensionality Reduction Techniques for Visual Cluster Analysis: An Empirical Study

Dimensionality Reduction (DR) techniques can generate 2D projections and enable visual exploration of cluster structures of high-dimensional datasets. However, different DR techniques would yield various patterns, which significantly affect the performance of visual cluster analysis tasks. We present the results of a user study that investigates the influence of different DR techniques on visual cluster analysis. Our study focuses on the most concerned property types, namely the linearity and locality, and evaluates twelve representative DR techniques that cover the concerned properties. Four controlled experiments were conducted to evaluate how the DR techniques facilitate the tasks of 1) cluster identification, 2) membership identification, 3) distance comparison, and 4) density comparison, respectively. We also evaluated users' subjective preference of the DR techniques regarding the quality of projected clusters. The results show that: 1) Non-linear and Local techniques are preferred in cluster identification and membership identification; 2) Linear techniques perform better than non-linear techniques in density comparison; 3) UMAP (Uniform Manifold Approximation and Projection) and t-SNE (t-Distributed Stochastic Neighbor Embedding) perform the best in cluster identification and membership identification; 4) NMF (Nonnegative Matrix Factorization) has competitive performance in distance comparison; 5) t-SNLE (t-Distributed Stochastic Neighbor Linear Embedding) has competitive performance in density comparison.

TacticFlow: Visual Analytics of Ever-Changing Tactics in Racket Sports

Event sequence mining is often used to summarize patterns from hundreds of sequences but faces special challenges when handling racket sports data. In racket sports (e.g., tennis and badminton), a player hitting the ball is considered a multivariate event consisting of multiple attributes (e.g., hit technique and ball position). A rally (i.e., a series of consecutive hits beginning with one player serving the ball and ending with one player winning a point) thereby can be viewed as a multivariate event sequence. Mining frequent patterns and depicting how patterns change over time is instructive and meaningful to players who want to learn more short-term competitive strategies (i.e., tactics) that encompass multiple hits. However, players in racket sports usually change their tactics rapidly according to the opponent's reaction, resulting in ever-changing tactic progression. In this work, we introduce a tailored visualization system built on a novel multivariate sequence pattern mining algorithm to facilitate explorative identification and analysis of various tactics and tactic progression. The algorithm can mine multiple non-overlapping multivariate patterns from hundreds of sequences effectively. Based on the mined results, we propose a glyph-based Sankey diagram to visualize the ever-changing tactic progression and support interactive data exploration. Through two case studies with four domain experts in tennis and badminton, we demonstrate that our system can effectively obtain insights about tactic progression in most racket sports. We further discuss the strengths and the limitations of our system based on domain experts' feedback.

Where Can We Help? A Visual Analytics Approach to Diagnosing and Improving Semantic Segmentation of Movable Objects

Semantic segmentation is a critical component in autonomous driving and has to be thoroughly evaluated due to safety concerns. Deep neural network (DNN) based semantic segmentation models are widely used in autonomous driving. However, it is challenging to evaluate DNN-based models due to their black-box-like nature, and it is even more difficult to assess model performance for crucial objects, such as lost cargos and pedestrians, in autonomous driving applications. In this work, we propose VASS, a Visual Analytics approach to diagnosing and improving the accuracy and robustness of Semantic Segmentation models, especially for critical objects moving in various driving scenes. The key component of our approach is a context-aware spatial representation learning that extracts important spatial information of objects, such as position, size, and aspect ratio, with respect to given scene contexts. Based on this spatial representation, we first use it to create visual summarization to analyze models' performance. We then use it to guide the generation of adversarial examples to evaluate models' spatial robustness and obtain actionable insights. We demonstrate the effectiveness of VASS via two case studies of lost cargo detection and pedestrian detection in autonomous driving. For both cases, we show quantitative evaluation on the improvement of models' performance with actionable insights obtained from VASS.

Towards Visual Explainable Active Learning for Zero-Shot Classification

Zero-shot classification is a promising paradigm to solve an applicable problem when the training classes and test classes are disjoint. Achieving this usually needs experts to externalize their domain knowledge by manually specifying a class-attribute matrix to define which classes have which attributes. Designing a suitable class-attribute matrix is the key to the subsequent procedure, but this design process is tedious and trial-and-error with no guidance. This paper proposes a visual explainable active learning approach with its design and implementation called semantic navigator to solve the above problems. This approach promotes human-AI teaming with four actions (ask, explain, recommend, respond) in each interaction loop. The machine asks contrastive questions to guide humans in the thinking process of attributes. A novel visualization called semantic map explains the current status of the machine. Therefore analysts can better understand why the machine misclassifies objects. Moreover, the machine recommends the labels of classes for each attribute to ease the labeling burden. Finally, humans can steer the model by modifying the labels interactively, and the machine adjusts its recommendations. The visual explainable active learning approach improves humans' efficiency of building zero-shot classification models interactively, compared with the method without guidance. We justify our results with user studies using the standard benchmarks for zero-shot classification.

VisQA: X-raying Vision and Language Reasoning in Transformers

Visual Question Answering systems target answering open-ended textual questions given input images. They are a testbed for learning high-level reasoning with a primary use in HCI, for instance assistance for the visually impaired. Recent research has shown that state-of-the-art models tend to produce answers exploiting biases and shortcuts in the training data, and sometimes do not even look at the input image, instead of performing the required reasoning steps. We present VisQA, a visual analytics tool that explores this question of reasoning vs. bias exploitation. It exposes the key element of state-of-the-art neural models - attention maps in transformers. Our working hypothesis is that reasoning steps leading to model predictions are observable from attention distributions, which are particularly useful for visualization. The design process of VisQA was motivated by well-known bias examples from the fields of deep learning and vision-language reasoning and evaluated in two ways. First, as a result of a collaboration of three fields, machine learning, vision and language reasoning, and data analytics, the work lead to a better understanding of bias exploitation of neural models for VQA, which eventually resulted in an impact on its design and training through the proposition of a method for the transfer of reasoning patterns from an oracle model. Second, we also report on the design of VisQA, and a goal-oriented evaluation of VisQA targeting the analysis of a model decision process from multiple experts, providing evidence that it makes the inner workings of models accessible to users.

VideoModerator: A Risk-aware Framework for Multimodal Video Moderation in E-Commerce

Video moderation, which refers to remove deviant or explicit content from e-commerce livestreams, has become prevalent owing to social and engaging features. However, this task is tedious and time consuming due to the difficulties associated with watching and reviewing multimodal video content, including video frames and audio clips. To ensure effective video moderation, we propose VideoModerator, a risk-aware framework that seamlessly integrates human knowledge with machine insights. This framework incorporates a set of advanced machine learning models to extract the risk-aware features from multimodal video content and discover potentially deviant videos. Moreover, this framework introduces an interactive visualization interface with three views, namely, a video view, a frame view, and an audio view. In the video view, we adopt a segmented timeline and highlight high-risk periods that may contain deviant information. In the frame view, we present a novel visual summarization method that combines risk-aware features and video context to enable quick video navigation. In the audio view, we employ a storyline-based design to provide a multi-faceted overview which can be used to explore audio content. Furthermore, we report the usage of VideoModerator through a case scenario and conduct experiments and a controlled user study to validate its effectiveness.

Learn, Generate, Rank, Explain: A Case Study of Visual Explanation by Generative Machine Learning

While the computer vision problem of searching for activities in videos is usually addressed by using discriminative models, their decisions tend to be opaque and difficult for people to understand. We propose a case study of a novel machine learning approach for generative searching and ranking of motion capture activities with visual explanation. Instead of directly ranking videos in the database given a text query, our approach uses a variant of Generative Adversarial Networks (GANs) to generate exemplars based on the query and uses them to search for the activity of interest in a large database. Our model is able to achieve comparable results to its discriminative counterpart, while being able to dynamically generate visual explanations. In addition to our searching and ranking method, we present an explanation interface that enables the user to successfully explore the model’s explanations and its confidence by revealing query-based, model-generated motion capture clips that contributed to the model’s decision. Finally, we conducted a user study with 44 participants to show that by using our model and interface, participants benefit from a deeper understanding of the model’s conceptualization of the search query. We discovered that the XAI system yielded a comparable level of efficiency, accuracy, and user-machine synchronization as its black-box counterpart, if the user exhibited a high level of trust for AI explanation.

VADAF: Visualization for Abnormal Client Detection and Analysis in Federated Learning

Federated Learning (FL) provides a powerful solution to distributed machine learning on a large corpus of decentralized data. It ensures privacy and security by performing computation on devices (which we refer to as clients) based on local data to improve the shared global model. However, the inaccessibility of the data and the invisibility of the computation make it challenging to interpret and analyze the training process, especially to distinguish potential client anomalies. Identifying these anomalies can help experts diagnose and improve FL models. For this reason, we propose a visual analytics system, VADAF, to depict the training dynamics and facilitate analyzing potential client anomalies. Specifically, we design a visualization scheme that supports massive training dynamics in the FL environment. Moreover, we introduce an anomaly detection method to detect potential client anomalies, which are further analyzed based on both the client model’s visual and objective estimation. Three case studies have demonstrated the effectiveness of our system in understanding the FL training process and supporting abnormal client detection and analysis.