ALL KAIST GIST DGIST UNIST

전체 제목 학술대회명 키워드

기관

---

• kaist 57,812
• gist 9,252
• dgist 2,279
• unist 21,033

연도

---

• 2230 1
• 2206 1
• 2204 1
• 2033 1
• 2030 1
• 2026 12
• 2025 1,827
• 2024 3,817
• 2023 5,509
• 2022 5,887

분류

---

• 국내 35,763
• 국제 53,431

정렬

• 날짜순
• 제목순

• Budgeted Strong Community Search in Heterogeneous Graphs ACM SIGMOD International Conference on Management of Data, 인도, 2026-05-31 ~ 2026-06-05

• Budgeted Strong Community Search in Heterogeneous Graphs

  ---

  Zhang, Wentong ; Fang, Lanting ; Kim, Junghoon ; Zhang, Kaibo ; Kim, Dahee ; Yuan, Ye ; Wang, Guoren ; Feng, Kaiyu

  초록
  학술대회명	ACM SIGMOD International Conference on Management of Data
  개최국	인도
  개최장소
  개최일	2026-06-01
  저자 키워드
  DOI
  bsl.detail.conference.eqpmn.ko
  기관	UNIST

• Safe Trajectory Planning with Bernstein Polynomials Under Control Barrier Function Constraints American Control Conference (ACC), 미국, 2026-05-26 ~ 2026-05-29

• Safe Trajectory Planning with Bernstein Polynomials Under Control Barrier Function Constraints

  ---

  Lee, Jungeun ; Lee, Seongjae ; Jeon, Jeong hwan

  초록	—Ensuring safe trajectory planning of autonomous mobile robots in environments with static and dynamic obstacles is a critical challenge for real-world deployment. Existing approaches, such as nonlinear predictive control (NMPC) with control barrier functions (CBFs), can provide safety guarantees. However, their computational complexity scales with the prediction horizon, which can hinder real-world feasibility. We propose an NMPC algorithm that combines discrete-time CBFs with Bernstein polynomial parameterization. By constraining a constant number of Bernstein control points within a dynamically generated safe corridor, the entire trajectory is guaranteed to remain forward invariant within the safe set. It also significantly reduces computational complexity compared to conventional approaches. Simulation results in both static and dynamic obstacle environments demonstrate that it achieves a high success rate while maintaining similar safety margins to existing baselines. At the same time, computation time is reduced by between 60% and 90%, depending on the scenario, and it remains nearly constant even as the prediction horizon increases. These results demonstrate the efficiency, robustness, and scalability of the proposed approach for safe trajectory planning in diverse scenarios.
  학술대회명	American Control Conference (ACC)
  개최국	미국
  개최장소
  개최일	2026-05-26
  저자 키워드
  DOI
  bsl.detail.conference.eqpmn.ko
  기관	UNIST

• Generalizing deep reinforcement learning agents to unseen environments remains a significant challenge. One promising solution is Unsupervised Environment Design (UED), a co-evolutionary framework in which a teacher adaptively generates tasks with high learning potential, while a student learns a robust policy from this evolving curriculum. Existing UED methods typically measure learning potential via regret, the gap between optimal and current performance, approximated solely by value-function loss. Building on these approaches, we introduce the transition-prediction error as an additional term in our regret approximation. To capture how training on one task affects performance on others, we further propose a lightweight metric called Co-Learnability. By combining these two measures, we present Transition-aware Regret Approximation with Co-learnability for Environment Design (TRACED). Empirical evaluations show that TRACED produces curricula that improve zero-shot generalization over ICLR 2026 (The Fourteenth International Conference on Learning Representations), 브라질, 2026-04-23 ~ 2026-04-27

• Generalizing deep reinforcement learning agents to unseen environments remains a significant challenge. One promising solution is Unsupervised Environment Design (UED), a co-evolutionary framework in which a teacher adaptively generates tasks with high learning potential, while a student learns a robust policy from this evolving curriculum. Existing UED methods typically measure learning potential via regret, the gap between optimal and current performance, approximated solely by value-function loss. Building on these approaches, we introduce the transition-prediction error as an additional term in our regret approximation. To capture how training on one task affects performance on others, we further propose a lightweight metric called Co-Learnability. By combining these two measures, we present Transition-aware Regret Approximation with Co-learnability for Environment Design (TRACED). Empirical evaluations show that TRACED produces curricula that improve zero-shot generalization over

  ---

  Cho, Geonwoo ; Im, Jaegyun ; Lee, Jihwan ; Kim, Sejin ; Kim, Sundong

  초록	Generalizing deep reinforcement learning agents to unseen environments remains a significant challenge. One promising solution is Unsupervised Environment Design (UED), a co-evolutionary framework in which a teacher adaptively generates tasks with high learning potential, while a student learns a robust policy from this evolving curriculum. Existing UED methods typically measure learning potential via regret, the gap between optimal and current performance, approximated solely by value-function loss. Building on these approaches, we introduce the transition-prediction error as an additional term in our regret approximation. To capture how training on one task affects performance on others, we further propose a lightweight metric called Co-Learnability. By combining these two measures, we present Transition-aware Regret Approximation with Co-learnability for Environment Design (TRACED). Empirical evaluations show that TRACED produces curricula that improve zero-shot generalization over strong baselines across multiple benchmarks. Ablation studies confirm that the transition-prediction error drives rapid complexity ramp-up and that Co-Learnability delivers additional gains when paired with the transition-prediction error. These results demonstrate how refined regret approximation and explicit modeling of task relationships can be leveraged for sample-efficient curriculum design in UED.
  학술대회명	ICLR 2026 (The Fourteenth International Conference on Learning Representations)
  개최국	브라질
  개최장소	브라질
  개최일	2026-04-25
  저자 키워드
  DOI
  bsl.detail.conference.eqpmn.ko
  기관	GIST

• Moderately solvating electrolyte with fluorinated cosolvents for lean-electrolyte Li-S batteries 2025년 한국공업화학회 춘계학술대회, 대한민국, 2026-06-02 ~ 2025-06-05

• Moderately solvating electrolyte with fluorinated cosolvents for lean-electrolyte Li-S batteries

  ---

  Kim, Hee-Tak ; Kim, IlJu

  초록
  학술대회명	2025년 한국공업화학회 춘계학술대회
  개최국	대한민국
  개최장소	icc 제주
  개최일	2025-06-03
  저자 키워드
  DOI
  bsl.detail.conference.eqpmn.ko
  기관	KAIST

• Towards Better Utilization of Multiple Views for Bundle Recommendation CIKM '24: The 33rd ACM International Conference on Information and Knowledge Management, 미국, 2204-10-21 ~ 2024-10-25

• Towards Better Utilization of Multiple Views for Bundle Recommendation

  ---

  Kim, Kyungho ; Kim, Sunwoo ; Lee, Geon ; Shin, Kijung

  초록
  학술대회명	CIKM '24: The 33rd ACM International Conference on Information and Knowledge Management
  개최국	미국
  개최장소	Boise ID USA
  개최일	2024-10-21
  저자 키워드
  DOI	10.1145/3627673.3680003
  bsl.detail.conference.eqpmn.ko
  기관	KAIST

• Development of chemical probes for capturing protein arginine kinases 6th Asian Chemical Biology Conference 2023, 대한민국, 2230-82-01 ~ 2023-08-23

• Development of chemical probes for capturing protein arginine kinases

  ---

  Lee, Shin Hyeon ; Kee, Jung-Min

  초록	Protein kinases play a critical role in controlling numerous cellular regulation pathways and signal transduction.1 Protein arginine phosphorylation, one of the relatively acid-labile N- phosphorylations, has been identified in numerous prokaryotic and eukaryotic proteins.2 The only known protein Arg kinase is McsB in Gram-positive bacteria, which was characterized in 2009.3 McsB is a heat-stress responsive kinase and phosphorylates Arg residues in aggregated cytosolic proteins so that phosphoarginine (pArg) acts as a protein degradation tag recognized by the ClpC/ClpP protease system. Recently, bacterial-PROTAC was developed using pArg and ClpC- binding cyclic peptide to induce the degradation of the target protein.4 In contrast, no eukaryotic or mycobacterial protein Arg kinase has been identified despite the abundance of pArg sites in those organisms, hindering further studies of Arg phosphorylation. Herein, we report our progress in developing mechanism-based chemical probes to label and capture the elusive Arg kinase.
  학술대회명	6th Asian Chemical Biology Conference 2023
  개최국	대한민국
  개최장소
  개최일	2023-08-22
  저자 키워드
  DOI
  bsl.detail.conference.eqpmn.ko
  기관	UNIST

• Porous Structure via Catalytic Etching of Hexagonal Boron Nitride by Formation of Transition Metal Oxides and Their utilization for Catalyst The 14th International Conference on Surfaces, Coatings and Nanostructured Materials, 그리이스, 2030-71-00 ~ 2023-07-14

• Porous Structure via Catalytic Etching of Hexagonal Boron Nitride by Formation of Transition Metal Oxides and Their utilization for Catalyst

  ---

  Lee, Myeung-Jin ; Jeong, Bora ; Ye, Bora ; Cho, Seungho ; Kim, Hong-Dae

  초록
  학술대회명	The 14th International Conference on Surfaces, Coatings and Nanostructured Materials
  개최국	그리이스
  개최장소
  개최일	2023-07-11
  저자 키워드
  DOI
  bsl.detail.conference.eqpmn.ko
  기관	UNIST

• ARC 문제에서 그래프를 활용한 객체 탐지연구 2023 한국컴퓨터종합학술대회, 대한민국, 2033-06-17 ~ 2023-06-20

• ARC 문제에서 그래프를 활용한 객체 탐지연구

  ---

  Lim, Mintaek ; Hwang, Sanha ; Ualibekova, Sabina ; Kim, Sundong

  초록
  학술대회명	2023 한국컴퓨터종합학술대회
  개최국	대한민국
  개최장소	제주도
  개최일	2023-06-20
  저자 키워드
  DOI
  bsl.detail.conference.eqpmn.ko
  기관	GIST

• Domain Knowledge-Informed Functional Outlier Detection for Line Quality Control Systems 한국경영과학회-대한산업공학회 2022 춘계공동학술대회, 대한민국, 2206-01 ~ 2206-04

• Domain Knowledge-Informed Functional Outlier Detection for Line Quality Control Systems

  ---

  Mun, Jonghwan ; Yoon, Kwonin ; Kim, Gyeongjun ; Yoo, Jitae ; Kim, Heesun ; Ryu, Nayi ; Kim, Sungil

  초록
  학술대회명	한국경영과학회-대한산업공학회 2022 춘계공동학술대회
  개최국	대한민국
  개최장소
  개최일	2022-06-03
  저자 키워드
  DOI
  bsl.detail.conference.eqpmn.ko
  기관	UNIST

• 광 네트워크를 통한 실시간 미디어 전송을 위한 멀티캐스트 라우팅 기법에 관한 연구 SWCC2008, 대한민국, ? ~ ?

• 광 네트워크를 통한 실시간 미디어 전송을 위한 멀티캐스트 라우팅 기법에 관한 연구

  ---

  김종원

  초록
  학술대회명	SWCC2008
  개최국	대한민국
  개최장소
  개최일	?0?
  저자 키워드
  DOI
  bsl.detail.conference.eqpmn.ko
  기관	GIST