Machine unlearning, an emerging research topic focusing on data privacy compliance, enables trained models to erase information learned from specific data. While many existing methods indirectly address this issue by intentionally injecting incorrect supervision, they often result in drastic and unpredictable changes to decision boundaries and feature spaces, leading to training instability and undesired side effects. To address this challenge more fundamentally, we first analyze the changes in latent feature spaces between the original and retrained models, and observe that the feature representations of samples not included in training are closely aligned with the feature manifolds of previously seen samples. Building on this insight, we introduce a novel evaluation metric for machine unlearning, coined dimensional alignment, which measures the alignment between the eigenspaces of the forget and retain sets. We incorporate this metric as a regularizer loss to develop a robust and stable unlearning framework, which is further enhanced by a self-distillation loss and an alternating training scheme. Our framework effectively eliminates information from the forget set while preserving knowledge from the retain set. Finally, we identify critical flaws in existing evaluation metrics for machine unlearning and propose new tools that more accurately capture its fundamental objectives.

Image denoising approaches based on deep neural net-works often struggle with overfitting to specific noise distributions present in training data. This challenge per-sists in existing real-world denoising networks, which are trained using a limited spectrum of real noise distributions, and thus, show poor robustness to out-of-distribution real noise types. To alleviate this issue, we develop a novel training framework called Adversarial Frequency Mixup (AFM). AFM leverages mixup in the frequency domain to generate noisy images with distinctive and challenging noise characteristics, all the while preserving the properties of authentic real-world noise. Subsequently, incorporating these noisy images into the training pipeline enhances the denoising network's robustness to variations in noise distributions. Extensive experiments and analyses, con-ducted on a wide range of real noise benchmarks demon-strate that denoising networks trained with our proposed framework exhibit significant improvements in robustness to unseen noise distributions. The code is available at https://github.com/dhryougit/AFM.

A primary goal of class-incremental learning is to strike a balance between stability and plasticity, where models should be both stable enough to retain knowledge learned from previously seen classes, and plastic enough to learn concepts from new classes. While previous works demonstrate strong performance on class-incremental benchmarks, it is not clear whether their success comes from the models being stable, plastic, or a mixture of both. This paper aims to shed light on how effectively recent class-incremental learning algorithms address the stability-plasticity trade-off. We establish analytical tools that measure the stability and plasticity of feature representations, and employ such tools to investigate models trained with various algorithms on large-scale class-incremental benchmarks. Surprisingly, we find that the majority of class-incremental learning algorithms heavily favor stability over plasticity, to the extent that the feature extractor of a model trained on the initial set of classes is no less effective than that of the final incremental model. Our observations not only inspire two simple algorithms that highlight the importance of feature representation analysis, but also suggest that class-incremental learning approaches, in general, should strive for better feature representation learning.

Abstract Smart Home Technology (SHT) as assistive technology (AT) is becoming an important active research field in the field of rehabilitation. For this purpose, Home Modification (HM) is one of the most common ways to improve the quality of life of the Persons with physical disabilities (PwPD). In this context, we propose a new Smart Home-based Home Modification Program (SHbHM) to improve the quality of life for PwPD. Our method simply uses Bluetooth or extends to Wi-Fi and Zigbee networks. A pilot study was conducted with five PwPD at home to investigate the effectiveness of the program. The reported results show a high quality of life, and the occupational performance and satisfaction are greatly improved, indicating that it is an efficient alternative.

We tackle the task of image retrieval with text feedback, where a reference image and modifier text are combined to identify the desired target image. We focus on designing an image-text compositor, i.e., integrating multi-modal inputs to produce a representation similar to that of the target image. In our algorithm, Content-Style Modulation (CoSMo), we approach this challenge by introducing two modules based on deep neural networks: the content and style modulators. The content modulator performs local updates to the reference image feature after normalizing the style of the image, where a disentangled multi-modal non-local block is employed to achieve the desired content modifications. Then, the style modulator reintroduces global style information to the updated feature. We provide an in-depth view of our algorithm and its design choices, and show that it accomplishes outstanding performance on multiple image-text retrieval benchmarks. Our code can be found at: https://github.com/postBG/CosMo.pytorch

Deep neural networks are susceptible to learn biased models with entangled feature representations, which may lead to subpar performances on various downstream tasks. This is particularly true for under-represented classes, where a lack of diversity in the data exacerbates the tendency. This limitation has been addressed mostly in classification tasks, but there is little study on additional challenges that may appear in more complex dense prediction problems including semantic segmentation. To this end, we propose a model-agnostic and stochastic training scheme for semantic segmentation, which facilitates the learning of debiased and disentangled representations. For each class, we first extract class-specific information from the highly entangled feature map. Then, information related to a randomly sampled class is suppressed by a feature selection process in the feature space. By randomly eliminating certain class information in each training iteration, we effectively reduce feature dependencies among classes, and the model is able to learn more debiased and disentangled feature representations. Models trained with our approach demonstrate strong results on multiple semantic segmentation benchmarks, with especially notable performance gains on under-represented classes.

For a self-sustaining wireless communication system in the Internet-of-Things (IoT) networks, energy harvesting (EH) can be implemented at each user node as a constant renewable power supply source. Hence, an investigation into the use of wireless-powered communication network (WPCN) protocols to facilitate communication between an access point (AP) and multiple mobile users (MUs) is presented in this article. The AP has multiple antennas and operates in the full-duplex (FD) mode. The MUs, on the other hand, have single antennas and works in the half-duplex (HD) mode. Each MU communicating with the FD-AP is assigned to one of two groups, based on the time allocation and channel access for either uplink (UL) or downlink (DL) communication. The channel assignment, time resource, and power resource allocations are optimized to maximize the UL weighted sum rate. The sum-rate optimization problem is found to be nonconvex. Therefore, an iterative algorithm is investigated to optimize the UL weighted sum rate of the proposed FD-WPCN system. Next, the proposed FD-WPCN algorithm is modified for HD-WPCN-enabled communication between the AP and multiple MUs. Extensive simulations are conducted to verify the proposed algorithm for FD-WPCN and compare its performance with the HD-WPCN counterpart. From the simulation results, FD-WPCN outperformed HD-WPCN at a low AP transmit signal-to-noise ratio (SNR) region. The opposite behavior is observed for high AP transmit SNR due to increasing residual self-interference at the FD-AP.

Recent works on domain adaptation exploit adversarial training to obtain domain-invariant feature representations from the joint learning of feature extractor and domain discriminator networks. However, domain adversarial methods render suboptimal performances since they attempt to match the distributions among the domains without considering the task at hand. We propose Drop to Adapt (DTA), which leverages adversarial dropout to learn strongly discriminative features by enforcing the cluster assumption. Accordingly, we design objective functions to support robust domain adaptation. We demonstrate efficacy of the proposed method on various experiments and achieve consistent improvements in both image classification and semantic segmentation tasks. Our source code is available at https://github.com/postBG/DTA.pytorch.

In this paper, we tackle unsupervised domain adaptation, where a target domain is unlabeled and lies on a considerably different distribution from a source domain. To alleviate such data discrepancies, we coin a novel deep neural network architecture that consists of a classifier and a domain discriminator on top of a shared feature extractor. Toward efficient regularization, we delegate a generation of the adversarial attacks to the domain discriminator. We then leverage the domain adversarial images to let the classification network learn important semantic features across the domains. Specifically, we employ consistency loss function that enables the joint use of clean and adversarial data. We present extensive experimental results on various domain adaptation benchmarks to show the efficacy of the proposed method.

People with Spinal cord injuries are having difficulty in health care, and complications cause physical, social and economic losses. In severe cases, complications lead to death and require systematic management. In this study, ICT-based health care service was developed to manage the respiratory function and urinary function of the people with spinal cord injuries and to help adapt to daily living activities and social participation through home visit occupational therapy. A pilot study was conducted with five clients with spinal cord injuries to investigate the effectiveness of the intervention services. As a result, it was confirmed that satisfaction, importance, and difficulty were appropriate. In the future, RCT clinical studies will be needed to diversify intervention services and expand the number of patients.

Code dissemination is a main component of reprogramming which enables over-the-air software update in wireless sensor networks (WSNs). In this paper, we present an adaptive code dissemination based on link quality (ACODI), which aims to minimize energy consumption. Compared to prior works on code dissemination, ACODI has a variety of notable features. First, it dynamically adapts the payload size in terms of energy efficiency. Second, it provides very low-overhead link estimation method. Finally, it is gracefully integrated into Deluge, which is the de facto standard code dissemination protocol in WSNs, and implemented on the TinyOS platform with very small overhead in terms of computation and memory. Our experiments using TelosB motes in the indoor testbed show that ACODI outperforms Deluge-22 and Deluge-108, which are Deluge with a fixed payload size of 22 and 108 bytes, respectively, in terms of energy efficiency and completion time.

In IEEE 802.11 WLANs, a fixed low transmission rate is used for multicast transmissions regardless of channel conditions of receivers. This may lead to inefficient use of wireless channel, especially when access point (AP) has both unicast and multicast data frames in the transmission queue. In this paper, we propose a dual queue based rate selecting schedule scheme for efficient use of wireless channel. In contrast to the legacy WLAN AP having a single FIFO queue, AP in the proposed scheme maintains two separate queues at the AP, one for unicast data and another for multicast data. When the AP accesses wireless channel, it decides a type of data (unicast or multicast) and transmission rate to be sent according to the channel condition and the delay boundary of multicast data. Our simulation results show that the proposed scheme increases the average data rate and energy consumption saving up to 13.7% compared to the auto rate fallback (ARF) which determines the transmission data rate only considering the channel condition.