Search

Temporal adaptive convolution has demonstrated superior performance over static convolution techniques in video understanding. However, it needs to be improved in long-time series modeling and multi-scale feature-map adaptation. To address these challenges, we introduce spatio-temporal hybrid adaptive convolution (STHAC), designed to enhance the spatio-temporal modeling capabilities of convolution. This is achieved by learning a set of spatio-temporal calibration filters to mitigate the spatial invariance intrinsic to static convolution methods. Specifically, STHAC learns a linear combination of N adaptive filters by parallelizing two lightweight attention branches. The resulting linearly mixed filters incorporate spatial multi-scale prior knowledge and long-range temporal dependencies. These spatio-temporal calibration filters modulate each frame's static convolutional weight parameters, thereby endowing static convolution with spatial multi-scale adaptability and long-range temporal modeling capabilities. Compared to other dynamic convolution methods, our proposed calibration filters require fewer parameters and incur lower computational complexity. Moreover, we introduce an Omni-dimensional aggregation module to augment the spatio-temporal modeling capacity of STHAC. When combined with STHAC, this aggregation module forms the spatio-temporal adaptive module (STAM) that can replace static convolution. We implement a spatio-temporal dynamic network based on STAM to validate our approach. Experimental results indicate that our model is competitive with state-of-the-art convolutional neural network architectures on action recognition benchmarks such as Kinetics-400(K400) and Something-Something V2(SSV2). [ABSTRACT FROM AUTHOR]

[Display omitted] In recent times, digital technology has played a pivotal role in advancing postgraduate orthodontic education. This article comprehensively overviews digital technology's various applications, benefits, and drawbacks in postgraduate orthodontic education. Further, it examines its transformative impact on students. The primary focus of this review is on reshaping traditional teaching methods and learning experiences through innovative tools and techniques. Beginning with online learning platforms, virtual classrooms, and webinars, a comprehensive review of their role in facilitating remote education and creating interactive learning environments is provided. It also explores the integration of interactive content and multimedia in post-graduate education and emphasizes their ability to enhance students' engagement and comprehension. Augmented and virtual reality (AR/VR) technologies will also be explored, with emphasis on their role in revolutionizing practical learning experiences within postgraduate curricula. Additionally, the article discusses the part that adaptive learning and artificial intelligence (AI) play in utilizing data analytics to personalize education and optimize learning outcomes. The influence of social media platforms, mobile applications, digital libraries, and open educational resources (OER) on collaborative learning environments will be reviewed, showcasing their impact on fostering collaboration and accessibility among postgraduate students. Collaborative tools, simulations, and e-labs' role in enhancing hands-on learning experiences are highlighted. Finally, a critical analysis of the benefits and drawbacks of integrating digital technology in orthodontic education, addressing ethical and legal concerns arising from its implementation, will also be overviewed. In summary, this article explores a nuanced understanding of the transformative potential of digital technology in postgraduate orthodontic education. It encompasses various applications, benefits, challenges, and ethical considerations, guiding stakeholders toward informed integration strategies for these technologies in higher education. [ABSTRACT FROM AUTHOR]

Adaptive learning aims to tailor the learning experience, including content, navigation, presentation, and strategies, based on learners' cognitive and affective factors. However, many existing adaptive learning systems often fail to meet the diverse needs and preferences of learners, typically relying on unique factors such as learning style. To address this, a robust framework is proposed, incorporating an artificial intelligence (AI) driven adaptive learning model capable of considering multiple factors, including past performance, hobbies, and learning style. The approach leverages the k-means clustering algorithm to group learners with similar leisure interests and incorporates Support Vector Regression to predict student performance, utilizing demographic data and past performance as metrics. Furthermore, the implementation of the Personalized Study Guide plugin, based on k-means clustering, enables the system to create personalized learning paths tailored to individual learning styles. By adopting this approach, the objective is to enhance learner engagement and performance through the development of a robust adaptive learning system. [ABSTRACT FROM AUTHOR]

Copyright of Arts for Linguistic & Literary Studies is the property of Thamar University and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

In response to the current situation of poor adaptive learning performance in Agent production and sales negotiation and dynamic changes in negotiation environment, the author proposes a method based on machine learning algorithms. Consider the impact of conflict level, cooperation possibility, and negotiation remaining time on negotiations in a dynamic negotiation environment, and use the entropy method to determine the weights of three influencing factors and perform linear weighting. Based on the differences in current negotiation topics, a concession amplitude prediction model based on dynamic selective ensemble learning is constructed, and an optimization strategy for supply chain production and sales negotiation is proposed. The experimental results indicate that, in the adaptive negotiation strategy of a regular SVM single learning machine, the joint utility of the most successfully negotiated agents falls within the interval [0.55, 0.70], while the author's ensemble learning strategy mainly focuses on [0.6, 0.8], the author's strategy is relatively superior to ordinary learning strategies in terms of both the number of successfully negotiated agents and the joint utility. Compared with the single learning machine negotiation strategy, this strategy improves the success rate and joint utility of Agent adaptive learning, and ensures the benefits of both production and sales in the supply chain, achieving a mutually beneficial situation for both parties in cooperation. [ABSTRACT FROM AUTHOR]

Personalized learning, a pedagogical approach tailored to individual needs and capacities, has garnered considerable attention in the era of artificial intelligence (AI) and the fourth industrial revolution. This systematic literature review aims to identify key drivers of personalized learning and critically assess the role of AI in reinforcing these drivers. Following PRISMA guidelines, a thorough search was conducted across major peer-reviewed journal databases, resulting in the inclusion of 102 relevant studies published between 2013 and 2022. A combination of qualitative and quantitative analyses, employing categorization and frequency analysis techniques, was performed to discern patterns and insights from the literature. The findings of this review highlight several critical drivers that contribute to the effectiveness of personalized learning, both from a broad view of education and in the specific context of e-learning. Firstly, recognizing and accounting for individual student characteristics is foundational to tailoring educational experiences. Secondly, personalizing content delivery and instructional methods ensures that learning materials resonate with learners' preferences and aptitudes. Thirdly, customizing assessment and feedback mechanisms enables educators to provide timely and relevant guidance to learners. Additionally, tailoring user interfaces and learning environments fosters engagement and accessibility, catering to diverse learning styles and needs. Moreover, the integration of AI presents significant opportunities to enhance personalized learning. AI-driven solutions offer capabilities such as automated learner profiling, adaptive content recommendation, realtime assessment, and the development of intelligent user interfaces, thereby augmenting the personalization of learning experiences. However, the successful adoption of AI in personalized learning requires addressing various challenges, including the need to develop educators' competencies, refine theoretical frameworks, and navigate ethical considerations surrounding data privacy and bias. By providing a comprehensive understanding of the drivers and implications of AI-driven personalized learning, this review offers valuable insights for educators, researchers, and policymakers in the Education 4.0 era. Leveraging the transformative potential of AI while upholding robust pedagogical principles, personalized learning holds the promise of unlocking tailored educational experiences that maximize individual potential and relevance in the digital economy. [ABSTRACT FROM AUTHOR]

Categorical learning is important and often challenging in both specialized domains, such as medical image interpretation, and commonplace ones, such as face recognition. Research has shown that comparing items from different categories can enhance the learning of perceptual classifications, particularly when those categories appear highly similar. Here, we developed and tested novel adaptively triggered comparisons (ATCs), in which errors produced during interactive learning dynamically prompted the presentation of active comparison trials. In a facial identity paradigm, undergraduate participants learned to recognize and name varying views of 22 unknown people. In Experiment 1, single-item classification trials were compared to a condition in which ATC trials were generated whenever a participant repeatedly confused two faces. Comparison trials required discrimination between simultaneously presented exemplars from the confused categories. In Experiment 2, an ATC condition was compared to a non-adaptive comparison condition. Participants learned to accuracy and speed criteria, and completed immediate and delayed posttests. ATCs substantially enhanced learning efficiency in both experiments. These studies, using a novel adaptive procedure guided by each learner’s performance, show that adaptively triggered comparisons improve category learning. [ABSTRACT FROM AUTHOR]

This paper presents an interdisciplinary framework, Machine Psychology, which integrates principles from operant learning psychology with a particular Artificial Intelligence model, the Non-Axiomatic Reasoning System (NARS), to advance Artificial General Intelligence (AGI) research. Central to this framework is the assumption that adaptation is fundamental to both biological and artificial intelligence, and can be understood using operant conditioning principles. The study evaluates this approach through three operant learning tasks using OpenNARS for Applications (ONA): simple discrimination, changing contingencies, and conditional discrimination tasks. In the simple discrimination task, NARS demonstrated rapid learning, achieving 100% correct responses during training and testing phases. The changing contingencies task illustrated NARS's adaptability, as it successfully adjusted its behavior when task conditions were reversed. In the conditional discrimination task, NARS managed complex learning scenarios, achieving high accuracy by forming and utilizing complex hypotheses based on conditional cues. These results validate the use of operant conditioning as a framework for developing adaptive AGI systems. NARS's ability to function under conditions of insufficient knowledge and resources, combined with its sensorimotor reasoning capabilities, positions it as a robust model for AGI. The Machine Psychology framework, by implementing aspects of natural intelligence such as continuous learning and goal-driven behavior, provides a scalable and flexible approach for real-world applications. Future research should explore using enhanced NARS systems, more advanced tasks and applying this framework to diverse, complex tasks to further advance the development of human-level AI. [ABSTRACT FROM AUTHOR]

Objective: In this study, we examined whether vestibular migraine, as a source of increased perceptual uncertainty due to the associated dizziness, interferes with adaptive learning. Methods: The IOWA gambling task (IGT) was used to assess adaptive learning in both healthy controls and patients with migraine-related dizziness. Participants were presented with four decks of cards (A, B, C, and D) and requested to select a card over 100 trials. Participants received a monetary reward or a penalty with equal probability when they selected a card. Card decks A and B (high-risk decks) involved high rewards (win £100) and high penalties (lose £250), whereas C and D (low-risk decks; favorable reward-to-punishment ratio) involved lower rewards (win £50) and penalties (lose £50). Task success required participants to decide (i.e., adaptively learn) through the feedback they received that C and D were the advantageous decks. Results: The study revealed that patients with vestibular migraine selected more high-risk cards than the control group. Chronic vestibular migraine patients showed delayed improvement in task performance than those with acute presentation. Only in acute vestibular migraine patients, we observed that impaired learning positively correlated with measures of dizzy symptoms. Conclusion: The findings of this study have clinical implications for how vestibular migraine can affect behavioural adaption in patients, either directly through altered perception or indirectly by impacting cognitive processes that can result in maladaptive behavior. [ABSTRACT FROM AUTHOR]

The native forestry industry is being reconsidered, or actively phased out, across most parts of Australia. This kind of transition has happened elsewhere in the world but, in Australia, there is no clarity from governments about what those forestry areas will become, from an ecological or legal perspective. In this research, we investigate the ecological, climatic, and governance contexts for this transition away from native forestry, placing our results in the context of the Global Biodiversity Framework's push to restore 30% of the Earth's land, coasts and oceans. We demonstrate important gaps in government planning for the future of former native forestry coupes, arguing that “after forestry” many of these places will need active intervention to recover these ecosystems and address historic land degradation. This case study combines legal analysis with a review of the restoration literature to illustrate the importance of the governance framework within which restoration science is evolving. We demonstrate the complexity of governing restoration at large scales as the climate changes, as well as potential opportunities to reconcile fragmented governance arrangements, to ensure that this transition results in resilient forest ecosystems in currently degraded areas. [ABSTRACT FROM AUTHOR]

The aim of this paper is to discuss the application of contemporary AI tools and technologies in education. The paper will mainly focus on a close examination of relevant academic literature as well as the current situation and methods of AI implementation in different sectors, finally, in education. Particularly, it was crucial that there were identified the major directions of applying AI in learning with an emphasis on learners and educators being the basic aspects of AI use within modern educational approaches. Secondly, some benefits and drawbacks concerning AI implementations in education supported by experiences from industry leaders and specific satisfaction measures among existing users, are analyzed. The most significant part covers integrating findings that show facts about artificial intelligence being used in education for 2024 like top online learning statistics for 2024, top AI statistics for 2024, and best AI plagiarism checkers for teachers in 2024 respectively. This includes evaluating teacher or instructor preparation skills to employ artificial intelligence (AI) tools and technology as well as ways students can learn how to use them without misuse towards quality education improvement. [ABSTRACT FROM AUTHOR]

Using anthropological theory, this paper examines human–android interactions (HAI) as an emerging aspect of android science. These interactions are described in terms of adaptive learning (which is largely subconscious). This article is based on the observations reported and supplementary data from two studies that took place in Japan with a teleoperated android robot called Telenoid in the socialization of school children and older adults. We argue that interacting with androids brings about a special context, an interval, and a space/time for reflection and imagination that was not there before. During the interaction something happens. There is adaptive learning and as a result, both children and older adults accepted Telenoid, and the children and older adults accepted each other. Using frames of play and ritual, we make sense and 'capture' moments of adaptive learning, and the feedback that elicits a social response from all study participants that results in self-efficacy and socialization. While "ritual" refers to the application of what has been learned and "play" means that there are no obvious consequences of what has been learned. This analysis illuminates new understanding about the uncanny valley, cultural robotics and the therapeutic potential of HAI. This has implications for the acceptance of androids in 'socialized roles' and gives us insight into the subconscious adaptive learning processes that must take place within humans to accept androids into our society. This approach aims to provides a clearer conceptual basis and vocabulary for further research of android and humanoid development. [ABSTRACT FROM AUTHOR]

As students' affective states can differ between learning that occurs in traditional classrooms when compared with learning that takes place in extended reality and immersive learning environments, it is important to examine the role of affective computing. Therefore, this study aims to provide an overview regarding the use of affective computing in the context of augmented reality, virtual reality, and immersive learning environments and the metaverse. Hence, it examines and presents the state of the art through a review and bibliometric analysis of 188 documents of the existing literature from Scopus and Web of Science (WoS) from 2005 to 2023. The study follows the Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA) statement to identify and select relevant to the topic documents. In addition to the analysis of the existing literature, emerging topics and themes are identified and future research directions are presented. The significant role of affective computing within augmented reality and virtual reality environments arose. Their ability to offer engaging and interactive learning experiences while also being able to recognize, monitor, and respond to students' affective states and to consider their emotions, personalities, characteristics, knowledge, and preferences to provide adaptive and personalized learning was evident. Additionally, their potential to enrich educational activities, increase learning outcomes, and support special education emerged. When augmented reality, virtual reality, and immersive learning environments capitalize on affective computing, meaningful learning can occur and self-regulated learning can be promoted. [ABSTRACT FROM AUTHOR]

Adaptive learning is a teaching approach aiming to personalize the learning experience for each student. In Computer Science Education (CSE), Adaptive Learning Systems (ALS) can provide students with customized lessons, exercises, and assessments based on their previous knowledge, strengths, and weaknesses. Some literature reviews focus on adaptive learning techniques, but none specifically analyze the methods used in adaptive systems in CSE. Our overall objective is to identify the computing techniques that are implemented within ALS to provide adaptive computer education in higher education. We conducted a scoping review (SR) based on the Guidelines for performing Systematic Literature Reviews in Software Engineering, defining six research questions and a search string. We performed searches in four databases, retrieving 512 documents. After applying the selection criteria, we selected 29 works. The results present quantitative and qualitative data from those 29 studies. Most research has used adaptive learning in programming education (12 articles) and databases (9 papers). Nine articles used learning styles as the core of adaptation, with the Felder-Silverman model being used by seven of them. The most commonly used approach was the rule-based system approach built by experts within the proposal, which was used in 17.24% of the articles. Clustering, collaborative filtering, fuzzy logic, and K-NN were used in 10.34% of the proposals. The positive aspect of "delivering adaptive content" was pointed out in seven articles. As for negative aspects, the problem of lack of student engagement was mentioned in four articles. This study discusses the application of dynamic and static models based on learning styles (LS) and the challenges involved in their use. Additionally, another aspect pointed out is that it is crucial to carefully consider pedagogical aspects in learning systems and have mixed development teams. The study also discusses student motivation and the use of gamification elements as a strategy to engage students in electronic learning systems. Finally, the results also indicate that the level of student knowledge is typically represented by a student model (SM), and the growing trend is the use of hybrid SMs. The research identified the most studied areas in ALS in CSE, such as computer programming and databases. The most common techniques used in ALS were rule-based systems built by experts, fuzzy logic, and clustering. The most common feature of ALS identified was learning resources, and the majority of studies focused on investigating the impact of system accuracy. This research can help researchers design better adaptive learning systems and understand challenges that need to be addressed. [ABSTRACT FROM AUTHOR]

Deep learning-based approaches have gained popularity for many applications in recent years and have become the state-of-the-art method in machine learning applications. However, most deep learning applications manipulate static data sets that do not vary over time. This is not the case for several significant recent applications, particularly for online learning on data streams that require real-time adjustment to the latest data context. This paper presents an architecture derived from the theory of transfer learning utilizing Extreme Learning Machine (ELM) and Recurrent Neural Network (RNN) methods. Unlike most data stream applications, when receiving high volumes of data at a high velocity, the proposed method does not store and reuse previous observations. The proposed model uses a Deep Echo State Network (DeepESN) for the primary classification task to compensate for the lack of past data. More precisely, instead of the data itself, we propagate generative models so that historical training data that is not stored can be regenerated. Before the DeepESN block, a layer of pre-trained ELMs was used to have generic feature engineering. The proposed method was tested on well-known benchmark data sets, and it outperformed the state-of-the-art models by an accuracy improvement ranging from 0.02 to 3.3 percent. The impact of data set regeneration on the learning process was studied using different RNN approaches. [ABSTRACT FROM AUTHOR]

Digital Information and Communication Technologies act as partners in the educational context, making it more dynamic through Virtual Learning Environments (VLEs). The adaptive system is based on technological solutions/tools, which allow the customization of teaching processes according to the student's singularities. Therefore, we conducted a systematic literature review (SLR) to elucidate which educational performance indicators best guide adaptive learning in virtual learning environments. To this end, we adopted the principles of Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) as a systematic review protocol, and as operational support, we used the online tool Parsifal. The initial database search - IEEE, ACM, and Scopus - returned 276 articles. After filtering based on the protocol, 16 articles remained part of the analysis and discussion corpus. The RSL results indicate that most of the indicators used to guide activities are based on the correctness and error of the questions. This shows that there is still much to be implemented in learning adaptability in virtual environments; for a more holistic assessment of learning, it is necessary to consider an integrated set of these indicators and not just individualized analyses. [ABSTRACT FROM AUTHOR]

Adaptive learning technologies are closely related to learners' self‐regulatory processes in individual and collaborative learning. This study presents the outcomes of a systematic literature review of empirical evidence on adaptive learning environments to foster self‐regulation and shared regulation of learning in collaborative settings. We provide an overview of what and how adaptive technologies have been used to understand and promote self‐regulated learning in collaborative contexts. A search resulted in 38 papers being analysed. Specifically, we identified the seven main objectives (feedback and scaffolding, self‐regulatory skills and strategies, learning trajectories, collaborative learning processes, adaptation and regulation, self‐assessment, and help‐seeking behaviour) that the adaptive technology research has been focusing on. We also summarize the implications derived from the reviewed papers and frame them within seven thematic areas. Finally, this review stresses that future research should consider developing a converging theoretical framework that would enable concrete monitoring and support for self‐regulation and socially shared regulation of learning. Our findings set a baseline to support the adoption and proliferation of adaptive learning technology within self‐regulated learning research and development. Practitioner notesWhat is already known about this topic By providing personalized and learner‐centric adaptive learning environments (ADLEs), adaptive learning technology can support and foster self‐regulated learning (SRL) practices.It is possible to create a more student‐centred and effective learning environment by combining adaptive learning and collaborative learning.Socially shared regulatory activities can involve planning, monitoring, controlling and reflecting on a group's learning processes.What this paper adds Provides a systematic literature review of empirical evidence on ADLEs, SRL and socially shared regulation of learning (SSRL) in collaborative contexts.Summarizes the insights on (S)SRL through ADLEs in collaborative learning.Identifies challenges and opportunities for ADLEs to support (S)SRL in collaborative learning.Implications for practice and/or policy Learning analytics and educational technology researchers will be able to use the systematic review as a guide for future research.Learning analytics and educational technology practitioners will be able to use the systematic review as a summary of the field's current state. [ABSTRACT FROM AUTHOR]

A compact MOSFET-C floating/grounded meminductor emulator (MIE) model is presented for high operating frequency and low power operation. The proposed MIE uses only 22 MOSFETs and two capacitors. Its performance is theoretically analyzed and rigorously verified using the Cadence Virtuoso software and hardware prototypes. The proposed MIE operates appropriately for a wide range of frequencies up to 5 MHz with $590 \mu \text{W}$ power consumption at a 180nm CMOS technology node and manifests important signature properties. The MIE layout area in 180 nm CMOS technology is $13107.5 \mu \text{m}^{2}$. To analyze the effects of statistical variations in MIE elements, extensive Monte Carlo simulations have been performed to demonstrate the robustness of the proposed MIE. For experimental validation, hardware prototypes have been developed and tested successfully. An MIE-based adaptive learning neuromorphic circuit is presented to show that it can mimic the behavioral responses of amoeba under varying environments such as temperature. [ABSTRACT FROM AUTHOR]

In the advent of e-learning, understanding student engagement and reaction is crucial for improving the quality of education and enhancing the learning rate. With the advancement of computer vision technologies, there is a significant opportunity to analyze and interpret student reactions in a non-intrusive manner. This study proposes a novel framework employing Faster R-CNN integrated with DenseNet architecture for real-time detection of student facial reactions during e-learning sessions. The proposed method leverages the strengths of Faster R-CNN in generating high-quality region proposals for object detection tasks, coupled with the DenseNet's efficiency in feature propagation and reduction in the number of parameters, which is well-suited for processing the intricate patterns in facial expressions. Our approach begins with the application of Faster R-CNN to extract potential facial regions with high accuracy and reduced computational cost. The integration of DenseNet as a backbone for feature extraction within Faster R-CNN capitalizes on its densely connected convolutional networks, ensuring maximum information flow between layers in the network. By doing so, the system becomes exceptionally adept at recognizing subtle changes in facial features that indicate various student reactions, such as confusion, engagement, or boredom. We conducted a series of experiments using a diverse dataset of e-learning interactions, collected under various lighting conditions and involving multiple ethnicities to ensure robustness and generalizability. The model was trained and validated on this dataset, and the results demonstrate a significant improvement in detection rates of student reactions compared to existing methods. [ABSTRACT FROM AUTHOR]

Copyright of Modern Educational Technology is the property of Editorial Board of Modern Educational Technology, Tsinghua University and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

The age of pervasive computing has initiated a boom in the development of adaptive context-aware learning environments (ACALEs), i.e., systems that are capable of detecting a learner's context and providing adaptive learning services based on this context. Many of the existing educational systems were developed as standalone applications for specific or a small range of adaptive educational scenarios. It would be extremely helpful for developers and educators to have a unified framework that provides an infrastructure for the development of ACALEs. In this study, we propose Lightlore—an adaptation framework that enables the development of different types of ACELEs for a wide range of learning scenarios in formal and informal settings. We first used scenario-based design (SBD) as the design methodology for creating a conceptual model of Lightlore. Educational scenarios were adopted from the results of a previous literature review. We then developed a proof-of-concept implementation of Lightlore, with a hypermedia system for learning data structures that uses the adaptation service of Lightlore. This implementation is essentially an adaptation infrastructure and a programming API for creating new (or transforming existing) adaptive and context-aware educational services. It exploits the experience API (xAPI), a modern e-learning standard and learning record store, thus making coupling with existing learning environments easier. We expect that diverse types of users will benefit from using Lightlore, such as learners, educators, learning environment developers, and researchers on educational technologies. [ABSTRACT FROM AUTHOR]

Abstract Categorical learning is important and often challenging in both specialized domains, such as medical image interpretation, and commonplace ones, such as face recognition. Research has shown that comparing items from different categories can enhance the learning of perceptual classifications, particularly when those categories appear highly similar. Here, we developed and tested novel adaptively triggered comparisons (ATCs), in which errors produced during interactive learning dynamically prompted the presentation of active comparison trials. In a facial identity paradigm, undergraduate participants learned to recognize and name varying views of 22 unknown people. In Experiment 1, single-item classification trials were compared to a condition in which ATC trials were generated whenever a participant repeatedly confused two faces. Comparison trials required discrimination between simultaneously presented exemplars from the confused categories. In Experiment 2, an ATC condition was compared to a non-adaptive comparison condition. Participants learned to accuracy and speed criteria, and completed immediate and delayed posttests. ATCs substantially enhanced learning efficiency in both experiments. These studies, using a novel adaptive procedure guided by each learner’s performance, show that adaptively triggered comparisons improve category learning.