Kyeong-Rae Kim

KAIST · School of Electrical Engineering

See. Hear. Feel.Make sense.

Bringing sight, sound, and touch together.
Exploring the world through signal processing
and multimodal learning.

Explore my research Say hello
Kyeong-Rae Kim sitting in a seaside capsule train.
Winter, South Korea
A RESEARCH NOTE

Signals, together.

Different signals. Shared understanding.
A LITTLE MORE ABOUT MEBASED IN DAEJEON, SOUTH KOREA

01 / A LITTLE ABOUT ME

A few words,
in my own way.

Hi, I’m Kyeong-Rae. I’m curious about how different senses help us understand the same world—and what AI can learn when we bring their signals together.

I’m an undergraduate in the School of Electrical Engineering at KAIST, with a focus on signal processing, expecting to graduate in February 2027. I’m currently a research intern at AMI Lab, advised by Prof. Tae-Hyun Oh.

My starting point is the signal: what it carries, how it changes, and what another sense can add. I’m interested in fusing visual, acoustic, and tactile signals into useful representations. My current work focuses on spatial audio-visual reasoning and physical consistency.

Signal processingMultimodal fusionSpatial intelligencePhysics-aware AI

RECENTLY

Small updates.
Forward motion.

  1. FloorSAV explores spatial audio-visual reasoning with a shared 2D floormap. PREPRINT

  2. Our work on physics-aware deepfake detection is at Interspeech 2026.

  3. Received the KAIST URP Excellent Award — top 3 in EECS.

02 / SELECTED RESEARCH

Connecting the dots.

Connecting signals across senses.
Grounding them in the physical world.

02 / PHYSICAL CONSISTENCY
Real and deepfake video examples: distance–speech consistency provides evidence when lip-sync cues alone are insufficient. Beyond lip-sync.CHECK THE PHYSICS BEHIND THE SIGNAL.

Interspeech2026

Physics-Aware Deepfake Detection via Distance–Speech Consistency

Kyeongrae Kim, Sung-Bin Kim, Hyun-Bin Oh, Tae-Hyun Oh

Looking beyond lip-sync by checking whether a speaker’s distance agrees with the acoustic properties of their voice.

Physics-aware AIDeepfake detectionAcoustics
Paper Code Project

03 / LEARNING BY DOING

Research
in practice.

Different questions.
The same drive to understand.

View full CV

AUG 2025 — PRESENT CURRENT

KAIST AMI Lab

Research Intern · Prof. Tae-Hyun Oh

Investigating physics-aware deepfake detection through acoustic measures and speaker position. Participated in the 2026 Winter/Spring Undergraduate Research Program.

Published at Interspeech 2026

JUN 2024 — JUL 2025

KAIST Smart Sound System Lab

Research Intern · Prof. Jung-Woo Choi

Explored indoor scene inpainting conditioned on first-order ambisonic room impulse responses. Participated in the 2025 Winter/Spring Undergraduate Research Program.

KAIST URP Excellent Award

AUG 2025 — JAN 2026 COLLABORATION

Seeing the finer details.

Unpaired super-resolution for fluorescence microscopy

Researched a physics-aware approach to super-resolution for in-vivo mouse neuron imaging, in a commissioned collaboration with a Ph.D. researcher at KAIST’s Graduate School of Medical Science and Engineering.

04 / FOUNDATIONS

Education.

2022 — PRESENT

KAIST

B.S. in Electrical Engineering

Signal Processing Division

Expected February 2027 · GPA 3.64 / 4.30

05 / A LITTLE ENCOURAGEMENT

Recognition.

NOVEMBER 2025

KAIST URP Excellent Award

Top 3 in EECS

Undergraduate Research Program

GOOD QUESTIONS START CONVERSATIONS.

What can we
make sense of next?

kimkyeongrae@kaist.ac.kr