The implementation of Hyperdimensional Computing (HDC) on In-Memory Computing (IMC) architectures faces significant challenges due to the mismatch between high-dimensional vectors and IMC array sizes, leading to inefficient memory utilization and increased computation cycles. This paper presents MEMHD, a Memory-Efficient Multi-centroid HDC framework designed to address these challenges. MEMHD introduces a clustering-based initialization method and quantization-aware iterative learning for multi-centroid associative memory. Through these approaches and its overall architecture, MEMHD achieves a significant reduction in memory requirements while maintaining or improving classification accuracy. Our approach achieves full utilization of IMC arrays and enables one-shot (or few-shot) associative search. Experimental results demonstrate that MEMHD outperforms state-of-the-art binary HDC models, achieving up to 13.69% higher accuracy with the same memory usage, or 13.25x more memory efficiency at the same accuracy level. Moreover, MEMHD reduces computation cycles by up to 80x and array usage by up to 71x compared to baseline IMC mapping methods when mapped to 128x128 IMC arrays, while significantly improving energy and computation cycle efficiency.

This paper addresses two critical challenges in analog In-Memory Computing (IMC) systems that limit their scalability and deployability: the computational unreliability caused by stuck-at faults (SAFs) and the high compilation overhead of existing fault-mitigation algorithms, namely Fault-Free (FF). To overcome these limitations, we first propose a novel multi-bit weight representation technique, termed row-column hybrid grouping, which generalizes conventional column grouping by introducing redundancy across both rows and columns. This structural redundancy enhances fault tolerance and can be effectively combined with existing fault-mitigation solutions. Second, we design a compiler pipeline that reformulates the fault-aware weight decomposition problem as an Integer Linear Programming (ILP) task, enabling fast and scalable compilation through off-the-shelf solvers. Further acceleration is achieved through theoretical insights that identify fault patterns amenable to trivial solutions, significantly reducing computation. Experimental results on convolutional networks and small language models demonstrate the effectiveness of our approach, achieving up to 8%p improvement in accuracy, 150 × faster compilation, and 2 × energy efficiency gain compared to existing baselines.

The deployment of deep neural networks on edge devices is a challenging task due to the increasing complexity of state-of-the-art models, requiring efforts to reduce model size and inference latency. Recent studies explore models operating at diverse quantization settings to find the optimal point that balances computational efficiency and accuracy. Truncation, an effective approach for achieving lower bit precision mapping, enables a single model to adapt to various hardware platforms with little to no cost. However, formulating a training scheme for deep neural networks to withstand the associated errors introduced by truncation remains a challenge, as the current quantization-aware training schemes are not designed for the truncation process. We propose TruncQuant, a novel truncation-ready training scheme allowing flexible bit precision through bit-shifting in runtime. We achieve this by aligning TruncQuant with the output of the truncation process, demonstrating strong robustness across bit-width settings, and offering an easily implementable training scheme within existing quantization-aware frameworks. Our code is released at https://github.com/a2jinhee/TruncQuant.

Multi-bit quantization networks enable flexible deployment of deep neural networks by supporting multiple precision levels within a single model. However, existing approaches suffer from significant training overhead as full-dataset updates are repeated for each supported bit-width, resulting in a cost that scales linearly with the number of precisions. Additionally, extra fine-tuning stages are often required to support additional or intermediate precision options, further compounding the overall training burden. To address this issue, we propose two techniques that greatly reduce the training overhead without compromising model utility: (i) Weight bias correction enables shared batch normalization and eliminates the need for fine-tuning by neutralizing quantization-induced bias across bit-widths and aligning activation distributions; and (ii) Bit-wise coreset sampling strategy allows each child model to train on a compact, informative subset selected via gradient-based importance scores by exploiting the implicit knowledge transfer phenomenon. Experiments on CIFAR-10/100, TinyImageNet, and ImageNet-1K with both ResNet and ViT architectures demonstrate that our method achieves competitive or superior accuracy while reducing training time up to 7.88×.

Interaction is crucial for online self-regulated learning (OSRL) ability and learning outcomes. The absence of social interaction might lead to high dropout rates in online learning environments. Danmaku holds great potential to enhance online learning by fostering interaction. This study explores the motivations behind university students’ use of danmaku and its influence on their OSRL. Using a mixed methods approach through surveys and interviews with 100 university students from two universities, we found that danmaku promotes social interaction by fulfilling students’ information and entertainment needs. Additionally, engagement with danmaku supports self-regulated learning through reflection and responding strategies and enhances enjoyment by increasing self-efficacy in contributing to the online learning community. This study expands understanding of the role interactive tools like danmaku can play in enhancing social interaction and OSRL, and highlights the potential of danmaku to improve student engagement and reduce dropout rates for quality education.

Large language models (LLMs) have demonstrated remarkable in-context learning capabilities across diverse applications. In this work, we explore the effectiveness of LLMs for generating realistic synthetic tabular data, identifying key prompt design elements to optimize performance. We introduce EPIC, a novel approach that leverages balanced, grouped data samples and consistent formatting with unique variable mapping to guide LLMs in generating accurate synthetic data across all classes, even for imbalanced datasets. Evaluations on real-world datasets show that EPIC achieves state-of-the-art machine learning classification performance, significantly improving generation efficiency. These findings highlight the effectiveness of EPIC for synthetic tabular data generation, particularly in addressing class imbalance.

Multivariate time-series prediction is a common task, but it often becomes challenging due to missing data caused by unreliable sensors and other issues. In fact, inaccurate imputation of missing values can degrade the downstream prediction performance, so it may be better not to rely on the estimated values of missing data. Furthermore, observed data may contain noise, so denoising them can be helpful for the main task at hand. In response, we propose a novel approach that can automatically utilize the optimal combination of the observed and the estimated values to generate not only complete, but also noise-reduced data by our own gating mechanism. We evaluate our model on incomplete real-world time-series datasets and achieved state-of-the-art performance. Moreover, we present in-depth studies using a carefully designed, synthetic multivariate time-series dataset to verify the effectiveness of the proposed model. The ablation studies and the experimental analysis of the proposed gating mechanism show that it works as an effective denoising and imputation method for time-series classification tasks.

Shadow detection is one of the most challenging issues in computer vision. Inspired by the great success of the convolutional neural network (CNN) for the problem of image restoration, learned features have been widely adopted for shadow detection. However, most existing methods still suffer from ambiguities driven by black-colored objects, which are not actually shaded, as well as the background clutter. In this letter, we propose the attentive feedback feature pyramid network (AFFPN) for shadow detection in a single image. The key idea of the proposed method is to extract shadow-relevant features based on multiple feedback modules, which are defined in the feature pyramid network. Specifically, attentive features extracted from each level of the encoder are progressively refined via connections between feedback modules from high-level to low-level layers for learning properties of shadow more accurately. Experimental results on benchmark datasets show that the proposed method is effective for shadow detection under complicated real-world environments. The code and model are publicly available at: https://github.com/JinheeKIM94/AFFPN_release.

We provide a more optimized model for calculating credibility score of information in SNS. We premeditated two heuristics which using characteristics of the credibility score for each document: (1) Expertise and (2) unbiasedness. Also, we divide the users in SNS into three types: (1) Creator (2) Distributor, and (3) Follower. Our model is designed to calculate Expertise and Un-biasedness for three types of SNS users (Creator, Distributor, and Follower) by using logistic regression model. Our model not only reveals whether the information is 'accurate and unbiased', but also investigates the 'source, distribution channel, and audience' of the information. We expect our credibility scoring will give answers to the 'qualitative problem' our online world is currently facing.