{"id":1134853,"date":"2025-03-21T02:19:07","date_gmt":"2025-03-21T09:19:07","guid":{"rendered":"https:\/\/www.microsoft.com\/en-us\/research\/?post_type=msr-blog-post&p=1134853"},"modified":"2025-03-26T22:50:28","modified_gmt":"2025-03-27T05:50:28","slug":"academic-growth-cultural-immersion-and-global-connections-korean-interns-journeys-at-msra","status":"publish","type":"msr-blog-post","link":"https:\/\/www.microsoft.com\/en-us\/research\/articles\/academic-growth-cultural-immersion-and-global-connections-korean-interns-journeys-at-msra\/","title":{"rendered":"Academic growth, cultural immersion, and global connections: Korean interns\u2019 journeys at MSRA"},"content":{"rendered":"\n

At the intersection of academic growth, cultural immersion, and international collaboration, internships at Microsoft Research Asia provide students with more than just technical experience.<\/p>\n\n\n\n

This year, twenty-two interns from research-oriented universities in Korea joined MSRA research groups, including Intelligent Multimedia, Machine Learning, and Media Computing. Their work spanned a wide range of exciting projects, from exploring responsible AI and tackling issues of fairness and privacy to hands-on experience with large language models and software engineering. These projects not only bridged theory and practice but also highlighted MSRA as a unique environment where collaboration, personal growth, and cultural exchange converge to shape the future leaders in technology.<\/p>\n\n\n\n

This article takes a closer look at each of their stories, showcasing their diverse research and how MSRA\u2019s unique environment has helped them push boundaries, forge connections, and uncover new possibilities for their careers and futures.<\/p>\n\n\n\n

Soyeon Kim: Mentorship, cultural immersion, and personal growth<\/h3>\n\n\n\n

Soyeon Kim, a third-year Ph.D. student at KAIST, focuses on responsible AI research, particularly fairness, differential privacy, and AI hallucinations. Motivated by her previous collaboration with MSRA in Korea, she chose to intern at MSRA under the mentorship of Partner Research Manager Xing Xie. She saw this as a unique opportunity to advance her research and have a meaningful impact on AI systems designed to benefit society.<\/p>\n\n\n\n

\"a<\/figure>\n\n\n\n

At MSRA, Soyeon worked on a project aimed at preventing AI systems from generating misleading answers while ensuring adaptability to diverse social contexts. This project, particularly in the field of LLM alignment, expanded her perspective and enhanced her research skills. Collaborating with researchers, she gained new insights and learned to communicate ideas more effectively, reinforcing the importance of teamwork in driving innovation.<\/p>\n\n\n\n

Though initially concerned about cultural and language differences, Soyeon adapted quickly to MSRA\u2019s inclusive and welcoming environment. Her mentor\u2019s guidance during setbacks was invaluable, teaching her to embrace failure as part of the learning process. \u201cHis support helped me view challenges as opportunities,\u201d she reflected. Additionally, her supervisor at KAIST, Steven Euijong Whang, inspired her to explore research at the intersection of databases and AI systems, offering advice on both academic and personal matters.<\/p>\n\n\n\n

\"Soyeon
Soyeon presented at the conference<\/figcaption><\/figure>\n\n\n\n

Beyond her research, Soyeon immersed herself in Beijing\u2019s culture, learning Chinese, sampling the local cuisine, and enjoying the city\u2019s winter charm. Reflecting on her time at MSRA, Soyeon sees it as a journey that not only advanced her academic work but also deepened her appreciation for cultural exchange and cross-disciplinary collaboration. She encourages others to embrace such opportunities and connect with people from diverse backgrounds to enrich their experiences.<\/p>\n\n\n\n

Dongjun Lee: Thriving in a global collaborative environment<\/h3>\n\n\n\n

Dongjun Lee, a first-year master\u2019s student at KAIST, pursued an internship with research teams of MSRA\u2019s Vancouver lab and Beijing lab, with the goal of developing AI systems for complex tasks like software engineering and AI research. Specializing in LLMs and multi-agent systems, Dongjun focused his research on improving unit test generation in software engineering. <\/strong><\/p>\n\n\n\n

\"Dongjun<\/figure>\n\n\n\n

By applying adversarial games, he worked on improving the efficiency and bug detection capabilities of test script generation, an area that evolved through close collaboration with his mentors, Senior Researcher Changho Hwang and Senior Researcher Ziyue Yang. Weekly discussions and feedback helped Dongjun refine his research direction, making meaningful progress. Additionally, his advisor Kimin Lee, whose main research interest is decision making agents and safety of AI, had regular one-on-one meetings with him every week during Dongjun\u2019s internship. Kimin Lee\u2019s insights greatly supported Dongjun\u2019s research at MSRA and contributed significantly to the paper writing process<\/p>\n\n\n\n

At MSRA, Dongjun found an ideal environment to explore production -level AI systems, enriched by a dynamic, multicultural atmosphere. Compared to his previous internship, he was particularly impressed by the seamless integration of global talent at MSRA, where researchers from different backgrounds exchanged ideas freely. This inclusive environment broadened his understanding of AI research in a global context, allowing him to gain fresh insights from colleagues with varied academic and cultural experiences.<\/p>\n\n\n\n

\"Dongjun
Dongjun in MSRA 2024 Korea Day<\/figcaption><\/figure>\n\n\n\n

Reflecting on his internship, Dongjun describes it as a pivotal journey. The opportunity to leverage MSRA\u2019s research resources, tackle challenges, and collaborate in a diverse, global environment has shaped his research approach and better prepared him for a promising career in AI.<\/p>\n\n\n\n

Jinyoung Jun: Addressing real-world challenges<\/h3>\n\n\n\n

Jinyoung Jun, a sixth-year Ph.D. candidate at Korea University, specializes in monocular depth-related tasks. During his internship at MSRA with the Media Computing Group, he focused on improving training data by addressing hardware sensor defects, aiming to create more reliable datasets for real-world applications. This work helps bridge the gap between theoretical research and practical implementation, a challenge Jinyoung finds both rewarding and crucial.<\/p>\n\n\n\n

\"Jinyoung<\/figure>\n\n\n\n

Jinyoung\u2019s journey to MSRA was guided by his supervisor’s recommendation. Professor Chang-Su Kim has guided him in building on his strengths while overcoming weaknesses with clear and insightful guidance. \u201cBeyond research, I\u2019ve developed essential qualities as an engineer, refining both my technical and problem-solving skills,\u201d Jinyoung reflected.<\/p>\n\n\n\n

At MSRA, Jinyoung gained access to large-scale computing resources while also encountering more complex technical challenges. Through weekly discussions with his mentors, Partner Research Manager Yan Lu, Senior Researcher Jiahao Li, and Senior Researcher Lei Chu, he gradually grasped the underlying logic of these challenges and learned to tackle them step by step. \u201cTheir feedback was crucial in guiding my work and helping me make significant progress,\u201d Jinyoung shared.<\/p>\n\n\n\n

\"Jinyoung
Jinyoung in MSRA 2024 Korea Day<\/figcaption><\/figure>\n\n\n\n

Additionally, MSRA\u2019s international research environment posed another challenge\u2014communicating complex technical concepts in English often felt overwhelming. However, through repeated exchanges and practice, he gradually discovered the key to effective communication: using clear and concise language while preparing relevant technical terms in advance. He believes that these seemingly fundamental skills will be invaluable for his future career.<\/p>\n\n\n\n

Looking ahead, Jinyoung sees his MSRA internship as a transformative experience. The mentorship, theoretical resources, and collaborative environment gave him the confidence to overcome difficulties, preparing him to face future challenges in both research and life.<\/p>\n\n\n\n

CheSang Park: Building global connections through collaboration<\/h3>\n\n\n\n

CheSang Park, a fifth-year Ph.D. candidate at Seoul National University, deepened his expertise in reinforcement learning and robotics during his MSRA internship in the Multi-Modal Computing group. His research on dynamic manipulation and collaborative robots was a significant milestone, shaping both his professional development and cultural growth.<\/p>\n\n\n\n

Encouraged by his professor, Chongwoo Park, CheSang applied for the internship, recognizing it as a unique opportunity to collaborate with world-class researchers. At MSRA, he focused on enhancing robot movement efficiency using foundation models in embodied AI, addressing computational delays that hinder performance. Under the mentorship of Senior Research Manager Jianlong Fu, he refined his research approach, learning to challenge assumptions and optimize methodologies.<\/p>\n\n\n\n

\"CheSang
CheSang and his group members<\/figcaption><\/figure>\n\n\n\n

Beyond research, CheSang embraced life in Beijing, immersing himself in local customs and daily experiences. He quickly grew fond of afternoon naps, a habit he found both refreshing and surprisingly effective in boosting productivity\u2014something he hadn\u2019t experienced in his home country. This small yet meaningful cultural adaptation helped him feel more in sync with his surroundings.<\/p>\n\n\n\n

He also built strong relationships with colleagues, actively participating in team events that brought everyone closer. One particularly memorable moment was playing Santa Claus during the Christmas celebrations. Donning a red suit and white beard, he handed out gifts and shared laughter with his teammates, fully embracing the festive spirit. This experience not only added joy to the occasion but also deepened his sense of belonging within the MSRA community.<\/p>\n\n\n\n

\"CheSang
CheSang plays Santa Claus at the Christmas party<\/figcaption><\/figure>\n\n\n\n

Reflecting on his internship, CheSang emphasized the impact of working alongside seasoned professionals. \u201cAt my university, I mostly collaborated with fellow students. Here, I have had the chance to work with experts who bring fresh perspectives,\u201d he shared. For CheSang, MSRA was more than just an internship\u2014it was a platform for innovation, collaboration, and personal growth. This experience highlighted the power of international cooperation and teamwork in shaping the future of robotics and AI.<\/p>\n\n\n\n

Youngsik Yoon: Growth at MSRA and AAAI 2025 paper milestone<\/h3>\n\n\n\n

Youngsik Yoon, a second-year Ph.D. candidate at Pohang University of Science and Technology (POSTECH), joined MSRA\u2019s Spatial Intelligence Group for his first internship, specializing in hierarchical reinforcement learning and online learning.  The experience not only deepened his expertise in AI research but also broadened his academic and cultural horizons, making it a pivotal step in his growth.<\/p>\n\n\n\n

When talking about the opportunity to intern at MSRA, Youngsik credited his professor, Jungseul Ok, whose influence and support played a crucial role.  During Jungseul\u2019s time as a graduate student, he also did an internship at MSRA, and his mentor at that time was Principal Researcher Wei Chen, who is now Youngsik\u2019s mentor. Therefore, before coming to MSRA, Youngsik was able to have regular meetings with Wei Chen, which allowed him to adapt in advance to conducting meetings and communicating in English.<\/p>\n\n\n\n

\"poster\"<\/figure>\n\n\n\n

His growth was also greatly shaped by his mentor, Wei Chen, whose continuous guidance helped refine his research. \u201cWei Chen\u2019s feedback significantly improved my research, weekly meetings, both online and in-person, allowed me to effectively address challenges and enhance the quality of my work.\u201d he noted.<\/p>\n\n\n\n

Youngsik valued the opportunity to collaborate with leading researchers and refine his ideas in a resource-rich setting. At MSRA, he focused on developing algorithms that balance theoretical rigor with practical effectiveness, culminating in his first major academic milestone\u2014submitting a theory-focused paper to AAAI 2025. This experience challenged him to present abstract ideas persuasively and highlight contributions without relying solely on experimental results.<\/p>\n\n\n\n

\"a<\/figure>\n\n\n\n

Reflecting on his time at MSRA, Youngsik highlighted how submitting his AAAI 2025 paper and experiencing the balance of research freedom and real-world impact opened his mind to opportunities in both academia and industry. MSRA was a defining chapter in his journey of growth and discovery.<\/p>\n\n\n\n

Jihyun Lee: Blending scientific research and real-world projects<\/h3>\n\n\n\n

Jihyun Lee, a fifth-year Ph.D. candidate at Yonsei University, embarked on an internship at MSRA in the Media Computing Group, guided by mentors Partner Research Manager Yan Lu and Senior Researcher Jiahao Li. Specializing in neural speech\/audio coding, multi-modal audio synthesis, and speech\/audio foundation models, Jihyun aimed to deepen her expertise while gaining experience in an international research environment.<\/p>\n\n\n\n

\"a
Jihyun and her mentor<\/figcaption><\/figure>\n\n\n\n

Before joining MSRA, Jihyun\u2019s professor Hong-Goo Kang consistently provided valuable guidance and insightful feedback on her research direction. Through Prof. Kang\u2019s guidance, Jihyun developed her research skills and the mindset of a researcher, which also encouraged her to pursue an internship at MSRA for further growth and collaboration.<\/p>\n\n\n\n

She also gained valuable experience in previous research. As a project engineer and later project leader, she developed DNN-based frameworks for speech\/audio compression, focusing on achieving high reconstruction quality and efficiency. She also led a project on a fully implantable brain-to-X system for voice communication, where she conducted surveys and pre-processed brain signal data to guide the project\u2019s direction. These experiences laid a strong foundation for her work at MSRA, particularly in multi-modal audio research.<\/p>\n\n\n\n

At MSRA, Jihyun\u2019s research involved stepping beyond her comfort zone, transitioning from speech synthesis to video-guided audio generation, allowing her to collaborate across disciplines. She appreciated the freedom to explore ideas without external deliverables, making her time at MSRA both fulfilling and productive. The guidance from her mentor and the collaborative environment helped refine her approach and shape her research direction.<\/p>\n\n\n\n

\"Jihyun
Jihyun in MSRA 2024 Korea Day<\/figcaption><\/figure>\n\n\n\n

Jihyun sees her MSRA internship as a defining experience that bridged scientific research and real-world applications. Engaging in interdisciplinary projects, she gained practical insights and global collaboration skills that will shape her future in both academia and industry.<\/p>\n\n\n\n

Miran Lee, who oversees the program and academic collaborations with Korea as director of the Microsoft Research Accelerator, emphasizes the importance of this program:<\/p>\n\n\n\n

 “Bringing Korean students to MSRA is about more than research – it’s about broadening perspectives, building global connections, and preparing students for the future of AI and technology.”<\/em><\/p>\n\n\n\n

For each of them, this internship was a chapter of growth, both professionally and personally, where the boundaries of knowledge intertwined with the richness of cultural exchange. These stories speak to the transformative power of stepping into the unknown. Their time at MSRA wasn\u2019t just about solving technical problems \u2013 it was about discovering new possibilities for themselves and the world around them. In a time when the future is uncertain, it\u2019s these connections, these shared experiences, that offer hope and promise for a brighter, more interconnected world, driven by the passion and collaboration of the next generation.<\/p>\n\n\n\n

\"a
The Korean interns completed their five-month internship and took a group photo after the closing party.<\/figcaption><\/figure>\n","protected":false},"excerpt":{"rendered":"

At the intersection of academic growth, cultural immersion, and international collaboration, internships at Microsoft Research Asia provide students with more than just technical experience. This year, twenty-two interns from research-oriented universities in Korea joined MSRA research groups, including Intelligent Multimedia, Machine Learning, and Media Computing. Their work spanned a wide range of exciting projects, from […]<\/p>\n","protected":false},"author":34512,"featured_media":0,"template":"","meta":{"msr-url-field":"","msr-podcast-episode":"","msrModifiedDate":"","msrModifiedDateEnabled":false,"ep_exclude_from_search":false,"_classifai_error":"","msr-content-parent":0,"msr_hide_image_in_river":null,"footnotes":""},"research-area":[13556],"msr-locale":[268875],"msr-post-option":[],"class_list":["post-1134853","msr-blog-post","type-msr-blog-post","status-publish","hentry","msr-research-area-artificial-intelligence","msr-locale-en_us"],"msr_assoc_parent":[],"_links":{"self":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-blog-post\/1134853","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-blog-post"}],"about":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/types\/msr-blog-post"}],"author":[{"embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/users\/34512"}],"version-history":[{"count":5,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-blog-post\/1134853\/revisions"}],"predecessor-version":[{"id":1135359,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-blog-post\/1134853\/revisions\/1135359"}],"wp:attachment":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/media?parent=1134853"}],"wp:term":[{"taxonomy":"msr-research-area","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/research-area?post=1134853"},{"taxonomy":"msr-locale","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-locale?post=1134853"},{"taxonomy":"msr-post-option","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-post-option?post=1134853"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}