SKSeungwoo Kang

Seungwoo Kang

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article

Previous Organizations / Institutes

The organizations / institutes from the scholar's other articles in the past ten years


Seungwoo Kang has a total of 49 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    MutterMeter: Earable Sensing of Self-Talk through Hierarchical Acoustic-Linguistic Fusion toward Self-Talk Interventions

    by Euihyeok Lee, Seonghyeon Kim, SangHun Im, Heung-Seon Oh, Seungwoo Kang

    Self-talk is a meaningful psychological signal that reflects individuals' emotions, motivations, and cognitive states, yet remains difficult to capture in daily life. It often emerges spontaneously during moments of pressure, concentration, or repeated decision-making, making it a relevant target for behavioral and cognitive interventions such as Educational Self-Talk Intervention (ESTI). However, ESTI largely depends on expert observation or retrospective self-reports, which are limited in capturing fleeting self-talk patterns and hinder its scalability in real-world settings. To address this gap, we present MutterMeter, a mobile self-talk detection system that continuously analyzes audio from earable microphones and classifies utterances into Negative self-talk, Positive self-talk, and Others. Focusing on tennis as a high-pressure and cognitively demanding testbed, MutterMeter employs a hierarchical framework that integrates acoustic, linguistic, and contextual information while adaptively balancing accuracy and efficiency. Evaluated on a first-of-its-kind dataset we collected (34.5h, n=29), MutterMeter achieves a macro-averaged F 1 score of 0.815 across subjects, outperforming baselines such as LLM-based models and speech emotion recognition models. Moreover, its adaptive processing pipeline reduces unnecessary computation by finalizing confident cases early, while maintaining robust detection performance.

  • Open Access

    Enabling Automatic Self-Talk Detection via Earables

    by Euihyeok Lee, Seonghyeon Kim, SangHun Im, Heung-Seon Oh, Seungwoo Kang

    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.

  • Open Access

    DJ-Fam: Using Favorite Songs as a Catalyst for Fostering Communication between Parents and Young Adult Children Living Apart

    by Euihyeok Lee, Souneil Park, Jin Yu, Seungchul Lee, Seungwoo Kang

    This paper aims to foster social interaction between parents and young adult children living apart via music. Our approach transforms their music-listening moment into an opportunity to listen to others' favorite songs and enrich interaction in their daily lives. To this end, we designed and implemented DJ-Fam, a mobile application that enables parents and children to listen to their favorite songs and use them as conversation starters to foster parent-child interaction. From our deployment study with seven families over four weeks in South Korea, we show the potential of DJ-Fam to positively influence parent-child interaction and their mutual understanding and relationship. Specifically, DJ-Fam considerably increases the frequency of communication and diversifies the communication channels and topics, all of which are satisfactory to the participants.


  • Open Access

    A Hybrid Model for Vehicle Sideslip Angle Estimation Based on Attention Regression

    by Seonghyeon Kim, Seung-Han You, Seungwoo Kang

    Vehicle Active Control Systems (ACS) have been developed to advance driver convenience and safety. These systems require accurate vehicle states such as lateral and longitudinal acceleration, sideslip angle, and yaw rate. However, achieving the desired accuracy without dedicated costly sensors is challenging. As a result, various methods have been developed for vehicle state estimation. This study proposes a hybrid model to estimate critical vehicle states, i.e., sideslip angle and yaw rate, by integrating a two-degree-offreedom single-track model (bicycle model) and a self-attention-based regression model. The regression model dynamically estimates tire cornering stiffness, a key parameter in the bicycle model. Using the varying estimates of tire cornering stiffness, the bicycle model accurately derives the sideslip angle and yaw rate. A newloss function is presented for practical learning of the attention regression model. Moreover, two learning strategies are proposed to enhance the model accuracy when actual measurement data of vehicle sideslip angle and yaw rate are unavailable. Experimental results show that the proposed hybrid model achieves higher accuracy than existing alternative methods.


  • Open Access

    Towards Generating Authentic Human-Removed Pictures in Crowded Places Using a Few-Second Video

    by Juhwan Lee, Euihyeok Lee, Seungwoo Kang

    If we visit famous and iconic landmarks, we may want to take a photo of them. However, such sites are usually crowded, and taking photos with only landmarks without people could be challenging. This paper aims to automatically remove people in a picture and produce a natural image of the landmark alone. To this end, it presents Thanos, a system to generate authentic human-removed images in crowded places. It is designed to produce high-quality images with reasonable computation cost using short video clips of a few seconds. For this purpose, a multi-frame-based recovery region minimization method is proposed. The key idea is to aggregate information partially available from multiple image frames to minimize the area to be restored. The evaluation result presents that the proposed method outperforms alternatives; it shows lower Fréchet Inception Distance (FID) scores with comparable processing latency. It is also shown that the images by Thanos achieve a lower FID score than those of existing applications; Thanos’s score is 242.8, while those by Retouch-photos and Samsung object eraser are 249.4 and 271.2, respectively.

  • Open Access

    GrooveMeter: Enabling Music Engagement-aware Apps by Detecting Reactions to Daily Music Listening via Earable Sensing

    by Euihyeok Lee, Chulhong Min, Jaeseung Lee, Jin Yu, Seungwoo Kang

    We present GrooveMeter, a novel system that automatically detects vocal and motion reactions to music and supports music engagement-aware applications. We use smart earbuds as sensing devices, already widely used for music listening, and devise reaction detection techniques by leveraging an inertial measurement unit (IMU) and a microphone on earbuds. To explore reactions in daily music-listening situations, we collect the first-kind-of dataset containing 926-minute-long IMU and audio data with 30 participants. With the dataset, we discover unique challenges in detecting music-listening reactions and devise sophisticated processing pipelines to enable accurate and efficient detection. Our comprehensive evaluation shows GrooveMeter achieves the macro F1 scores of 0.89 for vocal reaction and 0.81 for motion reaction with leave-one-subject-out (LOSO) cross-validation (CV). More importantly, it shows higher accuracy and robustness compared to alternative methods. We also present the potential use cases.


  • A Review of Foreign Community Interventions to Improve Walking Practices

    by Hyo Young Lee, 동서대학교보건행정학과, Seungwoo Kang

    Purpose: This study identified cases of overseas interventions conducted to improve walking practice to explore walking interventions that can be applied to improve walking practice rates in Korean communities. Method: A total of 11 documents were selected according to database search and literature selection criteria and finally analyzed. We searched PubMed, Google, and Google Scholar for literature published after 2000 using a combination of search terms (walk, walking, community, community-based population) Result: There were eight population-level studies and three individual-level studies covering the entire community, and the subjects were residents with physically inactive adults, the elderly, and socioeconomically vulnerable groups in the community. There were three randomised trials, four long-term cohort studies longer than one year, three cohort studies less than one year, and one mixed study. In addition, intervention types included two studies that provided economic incentives, two studies used community participatory interventions, three studies provided campaigns and information, three studies provided education and consultation, and six studies created an environment and infrastructure. The most studies created the environment and infrastructure. Conclusion: In order to achieve the 2030 target of walking practice rate 50.0% in South Korea, it is necessary to consider various topics such as building good governance through community participation, whole-of-government approaches, and changes in the social and environmental conditions of local communities, along with policy-enabling factors that can create an active systems centered on a systems-based approach that combines interventions at the individual and national levels.

  • AI-driven Family Interaction Over Melded Space and Time

    by Bumsoo Kang, Seungwoo Kang, Inseok Hwang

    Computer-mediated interaction services connect people over a distance. However, we address that those people are often “locked in a frame”—which includes an interaction mode, a point in time, or a context of either person. We observe that such lock-ins make it difficult to shape the interaction to be mutually symmetric. In this article, we propose a semantic-equivalent melding of space and time to provide a new form of empathetic interaction. We present HomeMeld and MomentMeld, which aim to meld space and time by applying AI, respectively. HomeMeld provides a sense of living together to a family living apart with AI-driven autonomous robotic avatars. MomentMeld utilizes an ensemble of visual AI to create interaction topics by matching semantic-equivalent photos. In-the-wild experiments reveal that HomeMeld and MomentMeld open new possibilities for empathetic interaction. Finally, we introduce a new interaction service leveraging technical synergy of HomeMeld and MomentMeld.


  • Open Access

    Enabling Product Recognition and Tracking Based on Text Detection for Mobile Augmented Reality

    by Sangwon Hwang, Jisun Lee, Seungwoo Kang

    We propose a system that supports real-time product recognition and tracking based on text detection for mobile augmented reality. To accurately distinguish products with visually similar packages, we develop a method that recognizes product names by utilizing the characteristics of texts printed on the product packages. It first filters out irrelevant products and effectively ranks candidate products through an inverted index search. We significantly reduce processing overhead by selectively performing product name recognition. In addition, we present an optical-flow-based method that enables efficient and responsive product tracking. Our evaluation shows that the proposed system achieves significantly better product recognition accuracy (80%) compared to alternative solutions, Vuforia (55.4%) and MobileNetV2 (69.6%). We also show that it achieves reasonable tracking accuracy and processing latency to support quality mobile AR experiences.


  • Open Access

    Daydriex: Translating Nighttime Scenes towards Daytime Driving Experience at Night

    by Euihyeok Lee, Seungwoo Kang

    What if the window of our cars is a magic window, which transforms dark views outside of the window at night into bright ones as we can see in the daytime? To realize such a window, one of important requirements is that the stream of transformed images displayed on the window should be of high quality so that users perceive it as real scenes in the day. Although image-to-image translation techniques based on Generative Adversarial Networks (GANs) have been widely studied, night-to-day image translation is still a challenging task. In this paper, we propose Daydriex, a processing pipeline to generate enhanced daytime translation focusing on road views. Our key idea is to supplement the missing information in dark areas of input image frames by using existing daytime images corresponding to the input images from street view services. We present a detailed processing flow and address several issues to realize our idea. Our evaluation shows that the results by Daydriex achieves lower Fréchet Inception Distance (FID) scores and higher user perception scores compared to those by CycleGAN only.