Sun-Wook Choi
Research Subjects
Sun-Wook Choi has a total of 25 co-authors. Here are the top co-authors by number of shared publications:
Dong Han Song
Department of Electronics Engineering, Pusan National University, Busan 609-735, Korea
3 shared publications- Giseop Kim
DGIST
3 shared publications Ki-Han Kim
Department of Electronics Engineering, Pusan National University, Busan 609-735, Korea
3 shared publicationsTae-Hoon Yoon
Department of Electronics Engineering, Pusan National University, Busan 609-735, Korea
3 shared publicationsByeong Koo Kim
IT/Mobile Development Group, LG Display Co., Ltd., Gumi, Gyeongbuk 730-700, Korea
2 shared publicationsByung Wok Park
Department of Electronics Engineering, Pusan National University, Busan 609-735, Korea
2 shared publications
Latest Publications
- 10 September 2026
- Published by Springer Nature
- p. 659-676
Streaming Dense Voxel Representationsfor 3D Occupancy Prediction
- 10 September 2026
- Published by Springer Nature
- p. 659-676
Streaming Dense Voxel Representationsfor 3D Occupancy Prediction
- 9 October 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Robotics and Automation Letters
- Vol. 10 (12) , 12365-12372
LAMP: Implicit Language Map for Robot Navigation
Recent advances in vision-language models have made zero-shot navigation feasible, enabling robots to interpret and follow natural language instructions without requiring labeling. However, existing methods that explicitly store language vectors in grid or node-based maps struggle to scale to large environments due to excessive memory requirements and limited resolution for fine-grained planning. We introduce LAMP (Language Map), a novel neural language field-based navigation framework that learns a continuous, language-driven map and directly leverages it for fine-grained path generation. Unlike prior approaches, our method encodes language features as an implicit neural field rather than storing them explicitly at every location. By combining this implicit representation with a sparse graph, LAMP supports efficient coarse path planning and then performs gradient-based optimization in the learned field to refine poses near the goal. Our two-stage pipeline of coarse graph search followed by language-driven, gradient-guided optimization is the first application of an implicit language map for precise path generation. This refinement is particularly effective at selecting goal regions not directly observed by leveraging semantic similarities in the learned feature space. To further enhance robustness, we adopt a Bayesian framework that models embedding uncertainty via the von Mises–Fisher distribution, thereby improving generalization to unobserved regions. To scale to large environments, LAMP employs a graph sampling strategy that prioritizes spatial coverage and embedding confidence, retaining only the most informative nodes and substantially reducing computational overhead. Our experimental results, both in NVIDIA Isaac Sim and on a real multi-floor building, demonstrate that LAMP outperforms existing explicit methods in both memory efficiency and fine-grained goal-reaching accuracy, opening new possibilities for scalable, language-driven robot navigation.
- 9 October 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Robotics and Automation Letters
- Vol. 10 (12) , 12365-12372
LAMP: Implicit Language Map for Robot Navigation
Recent advances in vision-language models have made zero-shot navigation feasible, enabling robots to interpret and follow natural language instructions without requiring labeling. However, existing methods that explicitly store language vectors in grid or node-based maps struggle to scale to large environments due to excessive memory requirements and limited resolution for fine-grained planning. We introduce LAMP (Language Map), a novel neural language field-based navigation framework that learns a continuous, language-driven map and directly leverages it for fine-grained path generation. Unlike prior approaches, our method encodes language features as an implicit neural field rather than storing them explicitly at every location. By combining this implicit representation with a sparse graph, LAMP supports efficient coarse path planning and then performs gradient-based optimization in the learned field to refine poses near the goal. Our two-stage pipeline of coarse graph search followed by language-driven, gradient-guided optimization is the first application of an implicit language map for precise path generation. This refinement is particularly effective at selecting goal regions not directly observed by leveraging semantic similarities in the learned feature space. To further enhance robustness, we adopt a Bayesian framework that models embedding uncertainty via the von Mises–Fisher distribution, thereby improving generalization to unobserved regions. To scale to large environments, LAMP employs a graph sampling strategy that prioritizes spatial coverage and embedding confidence, retaining only the most informative nodes and substantially reducing computational overhead. Our experimental results, both in NVIDIA Isaac Sim and on a real multi-floor building, demonstrate that LAMP outperforms existing explicit methods in both memory efficiency and fine-grained goal-reaching accuracy, opening new possibilities for scalable, language-driven robot navigation.
- 3 October 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Robotics and Automation Letters
- Vol. 9 (11) , 10700-10707
Addressing Diverging Training Costs Using BEVRestore for High-Resolution Bird's Eye View Map Construction
Recent advancements in Bird's Eye View (BEV) fusion for map construction have demonstrated remarkable mapping of urban environments. However, their deep and bulky architecture incurs substantial amounts of backpropagation memory and computing latency. Consequently, the problem poses an unavoidable bottleneck in constructing high-resolution (HR) BEV maps, as their large-sized features cause significant increases in costs including GPU memory consumption and computing latency, named diverging training costs issue. Affected by the problem, most existing methods adopt low-resolution (LR) BEV and struggle to estimate the precise locations of urban scene components like road lanes, and sidewalks. As the imprecision leads to risky motion planning like collision avoidance, the diverging training costs issue has to be resolved. In this paper, we address the issue with our novel BEVRestore mechanism. Specifically, our proposed model encodes the features of each sensor to LR BEV space and restores them to HR space to establish a memory-efficient map constructor. To this end, we introduce the BEV restoration strategy, which restores aliasing, and blocky artifacts of the up-scaled BEV features, and narrows down the width of the labels. Our extensive experiments show that the proposed mechanism provides a plug-and-play, memory-efficient pipeline, enabling an HR map construction with a broad BEV scope. Our code will be publicly released.
- 3 October 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Robotics and Automation Letters
- Vol. 9 (11) , 10700-10707
Addressing Diverging Training Costs Using BEVRestore for High-Resolution Bird's Eye View Map Construction
Recent advancements in Bird's Eye View (BEV) fusion for map construction have demonstrated remarkable mapping of urban environments. However, their deep and bulky architecture incurs substantial amounts of backpropagation memory and computing latency. Consequently, the problem poses an unavoidable bottleneck in constructing high-resolution (HR) BEV maps, as their large-sized features cause significant increases in costs including GPU memory consumption and computing latency, named diverging training costs issue. Affected by the problem, most existing methods adopt low-resolution (LR) BEV and struggle to estimate the precise locations of urban scene components like road lanes, and sidewalks. As the imprecision leads to risky motion planning like collision avoidance, the diverging training costs issue has to be resolved. In this paper, we address the issue with our novel BEVRestore mechanism. Specifically, our proposed model encodes the features of each sensor to LR BEV space and restores them to HR space to establish a memory-efficient map constructor. To this end, we introduce the BEV restoration strategy, which restores aliasing, and blocky artifacts of the up-scaled BEV features, and narrows down the width of the labels. Our extensive experiments show that the proposed mechanism provides a plug-and-play, memory-efficient pipeline, enabling an HR map construction with a broad BEV scope. Our code will be publicly released.
- Open Access
- 28 August 2013
- Published by IOP Publishing in Japanese Journal of Applied Physics
- Vol. 52 (9R) , 094102
Formation of Polymer Networks for Fast In-Plane Switching of Liquid Crystals at Low Temperatures
We formed a polymer structure to enable fast in-plane switching of liquid crystals at low temperatures. The problem of the inevitable slow response at low temperatures was reduced by the formation of in-cell polymer networks in in-plane switching (IPS) cells. The electro-optic characteristics of polymer-networked IPS cells were measured at temperatures ranging from -10 to 20 °C. The turn-on and turn-off times of an IPS cell were reduced by 44.5 and 47.2% at -10 °C by the formation of polymer networks. We believe that the proposed technology can be applied to emerging display devices such as mobile phones and automotive displays that may be used at low temperatures.
- 28 August 2013
- Published by IOP Publishing in Japanese Journal of Applied Physics
- Vol. 52 (9R) , 094102
Open AccessFormation of Polymer Networks for Fast In-Plane Switching of Liquid Crystals at Low Temperatures
We formed a polymer structure to enable fast in-plane switching of liquid crystals at low temperatures. The problem of the inevitable slow response at low temperatures was reduced by the formation of in-cell polymer networks in in-plane switching (IPS) cells. The electro-optic characteristics of polymer-networked IPS cells were measured at temperatures ranging from -10 to 20 °C. The turn-on and turn-off times of an IPS cell were reduced by 44.5 and 47.2% at -10 °C by the formation of polymer networks. We believe that the proposed technology can be applied to emerging display devices such as mobile phones and automotive displays that may be used at low temperatures.
- Open Access
- 25 April 2013
- Published by Korean Journal of Optics and Photonics in Journal of the Optical Society of Korea
- Vol. 17 (2) , 200-204
Fast Fringe-field-switching Liquid Crystal Cell with a Protrusion Structure
- 25 April 2013
- Published by Korean Journal of Optics and Photonics in Journal of the Optical Society of Korea
- Vol. 17 (2) , 200-204
Open AccessFast Fringe-field-switching Liquid Crystal Cell with a Protrusion Structure
- Open Access
- 25 June 2012
- Published by Korean Journal of Optics and Photonics in Journal of the Optical Society of Korea
- Vol. 16 (2) , 166-169
High-transmittance Multi-domain Vertical Alignment Liquid Crystal Device with Protrusion Structure
- 25 June 2012
- Published by Korean Journal of Optics and Photonics in Journal of the Optical Society of Korea
- Vol. 16 (2) , 166-169
Open AccessHigh-transmittance Multi-domain Vertical Alignment Liquid Crystal Device with Protrusion Structure