SeongUk Park
From MIPAL
SeongUk Park (박성욱)
|
Research interests : Model Compression, Generative Adversarial Networks, Deep learning.
Education
-
Mar. 2017 ~ Present:
- M.S./Ph.D. Student in Graduate School of Convergence Science and Technology, Seoul National University, Korea.
-
Feb. 2017:
- B.S. in Electronics Engineering, SungKyunKwan University, Korea.
Publications
- International Conferences
-
[1]Inseop Chung, SeongUk Park, Jangho Kim, Nojun Kwak, "Feature-map-level Online Adversarial Knowledge Distillation", Thirty-seventh International Conference on Machine Learning (ICML2020), July 2020, Online
[2] SeongUk Park and Nojun Kwak, "Feature-level Ensemble Knowledge Distillation for Aggregating Knowledge from Multiple Networks", 24th European Conference on Artificial Intelligence (ECAI 2020), Santiago, Spain, June. 2020.
[3] Simyung Chang, SeongUk Park, John Yang and Nojun Kwak, "Sym-Parameterized Dynamic Inference for Mixed-Domain Image Translation, International Conference on Computer Vision (ICCV2019), Seoul, Korea, Oct. 2019.
[4] Jangho Kim, SeongUK Park and Nojun Kwak, "Paraphrasing Complex Network: Network Compression via Factor Transfer", Thirty-second Conference on Neural Information Processing Systems (NIPS2018), Montreal, Canada, Dec. 2018.
[5] Simyung Chang, John Yang, SeongUk Park and Nojun Kwak, "Broadcasting Convolutional Network for Visual Relational Reasoning", European Conference on Computer Vision (ECCV2018), Munich, Germany, Sep. 2018.
Open Archive
-
[1] SeongUk Park, Kiyoon Yoo and Nojun Kwak, "On the Orthogonality of Knowledge Distillation with Other Techniques: From an Ensemble Perspective", arXiv, Sep. 2021.
[2] YoungJoon Yoo, SeongUk Park, Junyoung Choi, Sangdoo Yun and Nojun Kwak, "Butterfly Effect: Bidirectional Control of Classification Performance by Small Additive Perturbation", arXiv, Nov. 2017.
Projects
- 파노라마 영상과 사각지대경보(BSD) 기능을 제공하는 카메라 기반의 측후방 통합안전지원 시스템 개발, 산업통상자원부 (2019. 3 ~ )
- 딥러닝 기반 객체 인식 기술성능 고도화, (주)삼성전자 (2017. 9 ~ 2018. 9)