Education

Undergraduate coursework in International Trade

  • Mar. 2018 – Feb. 2019Mokpo National University, Mokpo, Republic of Korea
  • GPA: 3.64 / 4.50

B.S. in Computer Science

  • Mar. 2021 (transferred) – Feb. 2024Korea National Open University, Seoul, Republic of Korea
  • GPA: 4.40 / 4.50

M.S. in Electrical and Computer Engineering

Research Interests

Multimodal Emotion Recognition in Conversation, Knowledge Distillation, Human–Computer Interaction (HCI), Model Compression and Efficient AI

Publications & Under Review (Journal)

* (Corresponding Author)

International Journals

  1. DeogHwa KIM, Deok-Hwan KIM*, “DualDistill: Self and Cross-modal Knowledge Distillation for Multimodal Emotion Recognition in Conversation,” ICT Express (Impact Factor = 4.6)
  2. DeogHwa KIM, Yu Il LEE, Da Hyun YOON, Byeong Jun KIM, Deok-Hwan KIM*, “Dual-Distillation Vision-Language Model for Multimodal Emotion Recognition in Conversation with Quantized Edge Deployment,” Applied Sciences, vol. 16(6), pp. 3103, March 2026. (Impact Factor = 2.9)

Domestic Journals

  1. Han WANG, DeogHwa KIM, Deok-Hwan KIM*, “Enhancing Speech Emotion Recognition with Hybrid Graph Neural Networks : A GCN-GAT Framework,” The Journal of Korean Institute of Next Generation Computing, vol. 20, pp. 7-20, August 2024.

Publications (Conference)

International Conferences

  1. DeogHwa KIM, Han Wang, Deok-Hwan KIM*, “Multimodal Emotion Recognition based on Feature-level fusion of Facial Expression-Audio Modalities,” International Conference on Next Generation Computing (ICNGC), pp. 187-190, November 2024. (Oral Session; in Clark, Philippines)
  2. DeogHwa KIM, Jikyu PARK, Deok-Hwan KIM*, “MERC-KDC: Multimodal Emotion Recognition in Conversation Via Knowledge Distillation and Contrastive Learning,” International Conference on Ubiquitous and Future Networks (ICUFN), pp. 480-485, July 2025. (Oral Session; in Lisbon, Portugal)

Domestic Conferences

  1. DeogHwa KIM, Deok-Hwan KIM*, “MediaPipe Hand model-based sign language video hand joint recognition and LSTM-based data learning,” Korean Institute of Next Generation Computing, pp. 99-102, April 2024. (Oral Session; in Chungju, Republic of Korea)
  2. Seong Jun KWON, Seo Young YUN, Yu Il LEE, DeogHwa KIM, Deok-Hwan KIM*, “Study on Multi-Modal Emotion Recognition in Edge Computing Systems,” Korean Institute of Information Scientists and Engineers, pp. 1686-1688, December 2024. (Poster Session; in Yeosu, Republic of Korea)
  3. HaeDam SHIN, DeogHwa KIM, Deok-Hwan KIM*, “Lightweight Deep Learning Model using Response-based Class-wise Adaptive Temperature Knowledge Distillation,” Korean Institute of Next Generation Computing, pp. 236-239, May 2025. (Oral Session; in Asan, Republic of Korea)
  4. Min KIM, DeogHwa KIM, Deok-Hwan KIM*, “Jetson-Based Facial Emotion Recognition using Edge-Cloud Offloading,” The Institute of Electronics and Information Engineers, pp. 1207-1211, November 2025. (Oral Session; in Gyeonggi-do, Republic of Korea)

Patents

  1. Deok-Hwan KIM, HyunJun JO, Jikyu PARK, DeogHwa KIM, Han WANG, “A Training and Inference Framework for a Transformer-Based CAN Intrusion Detection System,” December 2024. (Computer Program, C-2024-051153)
  2. Deok-Hwan KIM, Han WANG, DeogHwa KIM, “Computer Device for Speech Emotion Recognition with Hybrid Graph Neural Networks GCN-GAT Framework,” August 2025. (Domestic Patent Application, 10-2025-0113351)

Projects

National Research Foundation of Korea

  • Project: A Study on multimodal AI based Speech and Emotion Assistant Solution in Edge Computing Environment

Inha University Software-Centered University Support Program

  • Project: Knowledge Distillation and Multimodal Learning Model for Conversational Emotion Recognition

Teaching Experience

  • C++ Programming (Learning Assistant, Spring 2024)
  • C++ Programming (Learning Assistant, Fall 2024)
  • Capstone Design in Electrical Electronic (Learning Assistant, Spring 2025)

Awards

Academic Excellence Award

  • Korea National Open University (2024)

Lifelong Learning Award

  • Korea National Open University (2024)

Honorable Mention Award

  • Embedded Software Contest (2024, 2025)