SangHun Im

Euihyeok Lee, Seonghyeon Kim, SangHun Im et al.

Self-talk-an internal dialogue that can occur silently or be spoken aloud-plays a crucial role in emotional regulation, cognitive processing, and motivation, yet has remained largely invisible and unmeasurable in everyday life. In this paper, we present MutterMeter, a mobile system that automatically detects vocalized self-talk from audio captured by earable microphones in real-world settings. Detecting self-talk is technically challenging due to its diverse acoustic forms, semantic and grammatical incompleteness, and irregular occurrence patterns, which differ fundamentally from assumptions underlying conventional speech understanding models. To address these challenges, MutterMeter employs a hierarchical classification architecture that progressively integrates acoustic, linguistic, and contextual information through a sequential processing pipeline, adaptively balancing accuracy and computational efficiency. We build and evaluate MutterMeter using a first-of-its-kind dataset comprising 31.1 hours of audio collected from 25 participants. Experimental results demonstrate that MutterMeter achieves robust performance with a macro-averaged F1 score of 0.84, outperforming conventional approaches, including LLM-based and speech emotion recognition models.

Kyungmin Kim, SangHun Im, GiBaeg Kim et al.

Text augmentation (TA) is a critical technique for text classification, especially in few-shot settings. This paper introduces a novel LLM-based TA method, TARDiS, to address challenges inherent in the generation and alignment stages of two-stage TA methods. For the generation stage, we propose two generation processes, SEG and CEG, incorporating multiple class-specific prompts to enhance diversity and separability. For the alignment stage, we introduce a class adaptation (CA) method to ensure that generated examples align with their target classes through verification and modification. Experimental results demonstrate TARDiS's effectiveness, outperforming state-of-the-art LLM-based TA methods in various few-shot text classification tasks. An in-depth analysis confirms the detailed behaviors at each stage.

SangHun Im, Gibaeg Kim, Heung-Seon Oh et al.

Hierarchical text classification (HTC) is essential for various real applications. However, HTC models are challenging to develop because they often require processing a large volume of documents and labels with hierarchical taxonomy. Recent HTC models based on deep learning have attempted to incorporate hierarchy information into a model structure. Consequently, these models are challenging to implement when the model parameters increase for a large-scale hierarchy because the model structure depends on the hierarchy size. To solve this problem, we formulate HTC as a sub-hierarchy sequence generation to incorporate hierarchy information into a target label sequence instead of the model structure. Subsequently, we propose the Hierarchy DECoder (HiDEC), which decodes a text sequence into a sub-hierarchy sequence using recursive hierarchy decoding, classifying all parents at the same level into children at once. In addition, HiDEC is trained to use hierarchical path information from a root to each leaf in a sub-hierarchy composed of the labels of a target document via an attention mechanism and hierarchy-aware masking. HiDEC achieved state-of-the-art performance with significantly fewer model parameters than existing models on benchmark datasets, such as RCV1-v2, NYT, and EURLEX57K.

Sanghun Im, Hyoungsuk Jeon, Jinho Choi et al.

We present a secret key agreement (SKA) protocol for a multi-user time-division duplex system where a base-station (BS) with a large antenna array (LAA) shares secret keys with users in the presence of non-colluding eavesdroppers. In the system, when the BS transmits random sequences to legitimate users for sharing common randomness, the eavesdroppers can attempt the pilot contamination attack (PCA) in which each of eavesdroppers transmits its target user's training sequence in hopes of acquiring possible information leak by steering beam towards the eavesdropper. We show that there exists a crucial complementary relation between the received signal strengths at the eavesdropper and its target user. This relation tells us that the eavesdropper inevitably leaves a trace that enables us to devise a way of measuring the amount of information leakage to the eavesdropper even if PCA parameters are unknown. To this end, we derive an estimator for the channel gain from the BS to the eavesdropper and propose a rate-adaptation scheme for adjusting the length of secret key under the PCA. Extensive analysis and evaluations are carried out under various setups, which show that the proposed scheme adequately takes advantage of the LAA to establish the secret keys under the PCA.