SERA LAB.KU · CSE

Korea University · Computer Science & Engineering

Systems for Extended Reality and AI Lab

We build the systems layer behind extended reality and AI — real-time media pipelines, on-device inference, and the networks underneath them.

Explore our research Work with us
A SERA Lab member giving a thumbs up at the lab booth during ICT AI TECH DAY, with the NPUsper and BlockStream posters behind
ICT AI TECH DAY2026 · Moment 01
See every photo
Korea UniversitySeoul · Republic of Korea

News

Recent news.

2026.09Publication

NPUsper has been accepted to NeurIPS 2026! Congratulations to Hojeong Lee, Sihyeon Lee, and Sungwon Woo! Sihyeon and Sungwon carried out this work during their undergraduate research internships.

2026.08Community

Three master’s students joined the lab.

2026.07Publication

A new paper was accepted to ACM SIGCOMM 2026.

2026.07Community

Two research interns joined the lab.

2026.03Publication

A new paper was accepted to IEEE SECON 2026.

2026.01Community

Two research interns joined the lab.

2026.01Award

Goodsol received a Gold Award at Samsung HumanTech Paper Awards 2026.

2025.07Publication

Two papers were accepted to USENIX NSDI 2026.

2025.07Publication

A new paper was accepted to IEEE INFOCOM 2026.

2025.05Community

Two research interns joined the lab.

2025.05Publication

A new paper was accepted to ACM MobiSys 2025.

2025.03Milestone

SERA Lab officially launched in the Department of Computer Science and Engineering at Korea University.

Research focus

What we work on.

See all research areas
01↗

Extended Reality Systems

End-to-end systems for immersive media: from volumetric video pipelines to latency-aware delivery and interaction.

  • Volumetric video
  • Real-time media
  • QoE
02↗

AI at the Edge

Practical AI systems that fit the constraints of devices, accelerators, networks, and the people using them.

  • Edge AI
  • NPU systems
  • On-device intelligence
03↗

Multi-Agent Systems

We design collaborative AI agents that reason, coordinate, and act together to solve complex tasks across real systems.

  • Multi-agent systems
  • Agentic AI
  • Distributed coordination
04↗

Networked Systems

Networks that adapt to the traffic they carry: cross-layer scheduling, transport, and measurement for real-time and learning-driven workloads.

  • Wireless networks
  • Transport & scheduling
  • Network measurement
05↗

AI Infrastructure

Scheduling, memory, and resource management for the clusters that train and serve models — where the workload itself is learned.

  • GPU cluster scheduling
  • LLM serving
  • RL for systems

Selected work

Recent papers.

Work from the lab and our collaborators, at systems, networking, and machine learning venues.

Browse publications
2026

ConferenceNeurIPS

NPUsper: Eliminating Redundant Computation for Real-Time Whisper on Mobile NPUs

AcceptedRead paper
2026

ConferenceACM SIGCOMM

DeepSFU: Scalable Deepfake Detection for Video Conferencing

Read paper
2026

ConferenceIEEE INFOCOM

PAVE: Mitigating Non-Congestive Delay for Seamless Video Calls over NextG Mobile Networks

Read paper
2026

ConferenceUSENIX NSDI

eXpressSFU: Toward Super-Scalable Video Conferencing with SmartNICs

Read paper
2026

ConferenceUSENIX NSDI

QCON: Seamless QoE-Aware 5G Streaming via Multi-Connectivity

Read paper
2026

ConferenceIEEE SECON

DualEngine: A Thermal-Aware Vision Inference Framework via Mobile and Cloud Co-Execution

Read paper

Build with us

We are recruiting graduate and undergraduate researchers.

Join SERA