Text-to-image (T2I) generative models can create vivid, realistic images from textual descriptions. As these models proliferate, they expose new concerns about their ability to represent diverse demographic groups, propagate stereo-types, and efface minority populations. Despite growing attention to the "safe" and "responsible" design of artificial intelligence (AI), there is no established methodology to systematically measure and control representational harms in image generation. This paper introduces a novel frame-work to measure the representation of intersectional groups in images generated by T2I models by applying the Multi-Group Proportional Representation (MPR) metric. MPR evaluates the worst-case deviation of representation statistics across given population groups in images produced by a generative model, allowing for flexible and context-specific measurements based on user requirements. We also develop an algorithm to optimize T2I models for this metric. Through experiments, we demonstrate that MPR can effectively mea-sure representation statistics across multiple intersectional groups and, when used as a training objective, can guide models toward a more balanced generation across demo-graphic groups while maintaining generation quality.1

Most continual learning (CL) algorithms have focused on tackling the stability-plasticity dilemma, that is, the challenge of preventing the forgetting of past tasks while learning new ones. However, we argue that they have overlooked the impact of knowledge transfer when the training dataset of a certain task is biased — namely, when the dataset contains some spurious correlations that can overly influence the prediction rule of a model. In that case, how would the dataset bias of a certain task affect the prediction rules of a CL model for future or past tasks? In this work, we carefully design systematic experiments using three benchmark datasets to answer the question from our empirical findings. Specifically, we first show through two-task CL experiments that standard CL methods, which are oblivious of the dataset bias, can transfer bias from one task to another, both forward and backward. Moreover, we find out this transfer is exacerbated depending on whether the CL methods focus on stability or plasticity. We then present that the bias is also transferred and even accumulates in longer task sequences. Finally, we offer a standardized experimental setup and a simple, yet strong plug-in baseline method, dubbed as group-class Balanced Greedy Sampling (BGS), which are utilized for the development of more advanced bias-aware CL methods.

In Reinforcement Learning (RL), designing precise reward functions remains to be a challenge, particularly when aligning with human intent. Preference-based RL (PbRL) was introduced to address this problem by learning reward models from human feedback. However, existing PbRL methods have limitations as they often overlook the second-order preference that indicates the relative strength of preference. In this paper, we propose Listwise Reward Estimation (LiRE), a novel approach for offline PbRL that leverages second-order preference information by constructing a Ranked List of Trajectories (RLT), which can be efficiently built by using the same ternary feedback type as traditional methods. To validate the effectiveness of LiRE, we propose a new offline PbRL dataset that objectively reflects the effect of the estimated rewards. Our extensive experiments on the dataset demonstrate the superiority of LiRE, i.e., outperforming state-of-the-art baselines even with modest feedback budgets and enjoying robustness with respect to the number of feedbacks and feedback noise. Our code is available at https://github.com/chwoong/LiRE

The notion of algorithmic fairness has been actively explored from various aspects of fairness, such as counterfactual fairness (CF) and group fairness (GF). However, the exact relationship between CF and GF remains to be unclear, especially in image classification tasks; the reason is because we often cannot collect counterfactual samples regarding a sensitive attribute, essential for evaluating CF, from the existing images (e.g., a photo of the same person but with different secondary sex characteristics). In this paper, we construct new image datasets for evaluating CF by using a high-quality image editing method and carefully labeling with human annotators. Our datasets, CelebA-CF and LFW-CF, build upon the popular image GF benchmarks; hence, we can evaluate CF and GF simultaneously. We empirically observe that CF does not imply GF in image classification, whereas previous studies on tabular datasets observed the opposite. We theoretically show that it could be due to the existence of a latent attribute $G$ that is correlated with, but not caused by, the sensitive attribute (e.g., secondary sex characteristics are highly correlated with hair length). From this observation, we propose a simple baseline, Counterfactual Knowledge Distillation (CKD), to mitigate such correlation with the sensitive attributes. Extensive experimental results on CelebA-CF and LFW-CF demonstrate that CF-achieving models satisfy GF if we successfully reduce the reliance on $G$ (e.g., using CKD).

The concepts of sacred places and spaces appear throughout religions globally. Places such as churches, cathedrals, temples, mosques, synagogues and graveyards are given special meanings, both functionally and spiritually, and separated from the ordinary. Recently location-based augmented reality (AR) technologies and applications have become widespread, and this raises questions regarding how AR content relates to sacred places. In this study, we approached this complex topic by asking clergy of the Lutheran Christian Church (N=47) to reflect on associated phenomena. We approached the data via reflexive thematic analysis and uncovered tensions related to (1) connected vs detached from sacredness; (2) supporting the spiritual purpose of the space vs conflicting with it; and (3) maintaining tradition vs embracing innovation. Overall, our findings suggest that AR technologies and products impact sacred spaces on multiple levels, but currently there is no consensus among the clergy on the impact of these changes.

There are numerous benefits from regularly walking in nature, and today's mobile technologies have the potential to encourage people to do so. Past research has showed that gamified map-based apps and location-based games (LBGs) have the capability to incentivize people to go to nature areas in cities and beyond. In this study, we explored LBGs' potential to bring people to nature by conducting a field experiment with a new mechanic called Routes in the popular LBG Pokémon GO. Prior to the Route feature's launch, we created altogether 13 Routes of various lengths in both city and nature landscapes. We collected numerical in-game data of how many times each Route was walked and deployed a survey (n=67) for Pokémon GO players in the area where the Routes were made. The findings suggest that proximity to population concentrations and in-game rewards are key drivers of Route popularity. Players' motivators to choose nature Routes over urban Routes were limited to outside-the-game factors such as scenery, and overall in our experiment the urban Routes turned out to be more popular.

Recently, fairness-aware learning have become increasingly crucial, but most of those methods operate by assuming the availability of fully annotated demographic group labels. We emphasize that such assumption is unrealistic for real-world applications since group label annotations are expensive and can conflict with privacy issues. In this paper, we consider a more practical scenario, dubbed as Algorithmic Group Fairness with the Partially annotated Group labels (Fair-PG). We observe that the existing methods to achieve group fairness perform even worse than the vanilla training, which simply uses full data only with target labels, under Fair-PG. To address this problem, we propose a simple Confidence-based Group Label assignment (CGL) strategy that is readily applicable to any fairness-aware learning method. CGL utilizes an auxiliary group classifier to assign pseudo group labels, where random labels are assigned to low confident samples. We first theoretically show that our method design is better than the vanilla pseudo-labeling strategy in terms of fairness criteria. Then, we empirically show on several benchmark datasets that by combining CGL and the state-of-the-art fairness-aware in-processing methods, the target accuracies and the fairness metrics can be jointly improved compared to the baselines. Furthermore, we convincingly show that CGL enables to naturally augment the given group-labeled dataset with external target label-only datasets so that both accuracy and fairness can be improved. Code is available at https://github.com/naver-ai/cgl_fairness.

Recent studies have demonstrated that gradient matching-based dataset synthesis, or dataset condensation (DC), methods can achieve state-of-theart performance when applied to data-efficient learning tasks. However, in this study, we prove that the existing DC methods can perform worse than the random selection method when taskirrelevant information forms a significant part of the training dataset. We attribute this to the lack of participation of the contrastive signals between the classes resulting from the class-wise gradient matching strategy. To address this problem, we propose Dataset Condensation with Contrastive signals (DCC) by modifying the loss function to enable the DC methods to effectively capture the differences between classes. In addition, we analyze the new loss function in terms of training dynamics by tracking the kernel velocity. Furthermore, we introduce a bi-level warm-up strategy to stabilize the optimization. Our experimental results indicate that while the existing methods are ineffective for fine-grained image classification tasks, the proposed method can successfully generate informative synthetic datasets for the same tasks. Moreover, we demonstrate that the proposed method outperforms the baselines even on benchmark datasets such as SVHN, CIFAR-10, and CIFAR-100. Finally, we demonstrate the high applicability of the proposed method by applying it to continual learning tasks.

The design and development of playful wearable devices is a challenging and complicated problem. It entails not only multidisciplinary expertise but also a comprehensive understanding of player experience. There is a scarcity of evidence-based studies in current state-of-art literature that investigate general design practices and provide pragmatic design implications and suggestions based on solid user-centered research. To bridge the gap, we developed five experience prototypes based on the speculative design concepts from previous studies, and a Wizard of Oz experiment was conducted to elicit end users' feedback regarding general gaming experience as well as specific design themes in different gaming scenarios. The user experiment results were analyzed qualitatively following a rigorous thematic analysis, generating five major design implications as output. We believe this study will offer forward-looking insights to designers, developers and the research community, facilitating future work in this field.

In this pictorial, we depict our design process on gaming wearables starting from participatory design workshops to concept creation. Wearables possess strong qualities for gaming such as performativity, sociality and interactivity. However, it is an emergent field and there is a dearth of design knowledge especially when it comes to designing wearables for mainstream gaming platforms such as game consoles. Our aim is to explore this field elaborately with a research through design approach and also clearly exemplify how our design process progressed through different phases. Our results, apart from helping wearables designers to understand critical features for mainstream gaming, will also demonstrate the techniques and methods for extracting knowledge from PD workshops and incorporating it in a conceptual design phase.

Although intelligent reflecting surface (IRS) is regarded as a promising solution to enhance the efficiency of wireless energy transfer (WET), the acquiring of channel state information is a crucial challenge for the system in which a training sequence for channel estimation is sent by low-power Internet-of-Things (IoT) devices. In this article, an IRS-aided multidevice WET system is considered. To overcome the limitation in channel estimation, we propose a received power-based channel estimation scheme that can be easily implemented and scalable in wirelessly empowered IoT devices. Specifically, at every single time slot, each device measures the received power of a randomly generated radio-frequency signal and feeds it back to the transmitter. We formulate a channel estimation problem to use the history of received power measurements based on the maximum-likelihood estimation using the phase retrieval framework and temporal channel evolution model. Moreover, we propose an algorithm that can be employed to obtain the stationary solution for the channel estimation problem, which is based on the inexact block coordinate descent method. We also perform algorithm modification to deal with the special case in which the transmitter-IRS channel is available. The simulation results show that the performance of the proposed algorithm approaches the upper bound as the channel slowly changes, although the proposed channel estimation protocol requires only one scalar value feedback.

We propose a novel regularization-based continual learning method, dubbed as Adaptive Group Sparsity based Continual Learning (AGS-CL), using two group sparsity-based penalties. Our method selectively employs the two penalties when learning each neural network node based on its the importance, which is adaptively updated after learning each task. By utilizing the proximal gradient descent method, the exact sparsity and freezing of the model is guaranteed during the learning process, and thus, the learner explicitly controls the model capacity. Furthermore, as a critical detail, we re-initialize the weights associated with unimportant nodes after learning each task in order to facilitate efficient learning and prevent the negative transfer. Throughout the extensive experimental results, we show that our AGS-CL uses orders of magnitude less memory space for storing the regularization parameters, and it significantly outperforms several state-of-the-art baselines on representative benchmarks for both supervised and reinforcement learning.

Complex relationships among entities can be modeled very effectively using hypergraphs. Hypergraphs model real-world data by allowing a hyperedge to include two or more entities. Clustering of hypergraphs enables us to group the similar entities together. While most existing algorithms solely consider the connection structure of a hypergraph to solve the clustering problem, we can boost the clustering performance by considering various features associated with the entities as well as auxiliary relationships among the entities. Also, we can further improve the clustering performance if some of the labels are known and we incorporate them into a clustering model. In this paper, we propose a semi-supervised clustering framework for hypergraphs that is able to easily incorporate not only multiple relationships among the entities but also multiple attributes and content of the entities from diverse sources. Furthermore, by showing the close relationship between the hypergraph normalized cut and the weighted kernel K-Means, we also develop an efficient multilevel hypergraph clustering method which provides a good initialization with our semi-supervised multi-view clustering algorithm. Experimental results show that our algorithm is effective in detecting the ground-truth clusters and significantly outperforms other state-of-the-art methods.