Homepage: Kisang Park

Hi, I'm Kisang Park.

I'm a Ph.D. student at the Robot Intelligence Lab (RILAB), Korea University, advised by Prof. Sungjoon Choi.

My passion lies in developing robotic agents capable of performing cooperative tasks.
I am particularly interested in how robots can coordinate their motions and decisions to accomplish complex tasks, with a focus on motion patterns and physical interactions.

News

[Nov 2026] I will be attending CoRL holding in Austin, Texas.

[Sep 2026] Our paper Learning Unions of Convex Sets via Invertible Latent Decomposition for Path Planning has been accepted to CoRL 2026!

[Jul 2026] I will be attending RSS holding in Sydney, Australia.

[Jun 2026] Our paper Efficient Timing-Aware Planning via Detachable Task Modeling has been accepted to IROS 2026!

[Apr 2026] Our paper Natural Functional Gradient for Smooth Trajectory Optimization has been accepted to RSS 2026!

[Sep 2025] I joined the Robot Intelligence Lab (RILAB) at Korea University, under the supervision of Prof. Sungjoon Choi.

Publications

Learning Unions of Convex Sets via Invertible Latent Decomposition for Path Planning
Taerim Yoon, Dongho Kang, Kisang Park, Junha Cha, Stelian Coros, Sungjoon Choi
CoRL 2026

Collision-free path planning in cluttered, real-world environments relies on a representation of the collision-free space, and existing representations broadly fall into two categories, explicit and implicit representations. We bridge this gap with ILD (Invertible Latent Decomposition), a framework that jointly learns an invertible mapping and a union of explicit convex polytopes in the resulting latent space.

Project page / ArXiv

Efficient Timing-Aware Planning via Detachable Task Modeling
Seungmin Kim, Kisang Park, Ina Jung, Hyeonseong Kim, Dongkyu Park, Yisoo Lee, Inho Lee, Yoonseon Oh,Kyeongjae Lee, Sungjoon Choi
IROS 2026

We introduce the concept of detachable tasks to capture such waiting intervals, enabling them to be separated from active execution and interleaved with other tasks. Building on this concept, we propose Timing-Aware Detachable-task planning (TAD) that transforms an off-the-shelf task plan into a timing-aware plan with explicit start and end times, enabling efficient single-robot execution by overlapping compatible tasks during waiting intervals.

Project page / Paper / ArXiv

Natural Functional Gradients for Smooth Trajectory Optimization
Kisang Park, Chanwoo Kim, Kyeongjae Lee, Sungjoon Choi
RSS 2026

We propose a trajectory optimization framework that performs geometry-aware updates directly in function space using natural functional gradients. The method optimizes a Gaussian-smoothed surrogate objective that regularizes the optimization landscape through smooth trajectory perturbations while preserving trajectory-level structure.

Project page / Paper / ArXiv