Hyoungseob Park
Papers
1
Total Citations
48
H-Index
1
About
Hyoungseob Park is a rising force in embodied AI and multimodal machine learning, whose work is redefining how machines perceive the physical world through touch. His primary research focuses on developing unified representations that bridge tactile sensing with vision, language, and other modalities—a critical step toward robots that can interact with objects as adeptly as humans. In his landmark 2024 paper, “Binding Touch to Everything,” Park introduced a framework for learning cross-modal tactile representations that generalize across diverse sensor types, overcoming the longstanding challenge of sparse, sensor-specific data. This work, already garnering 48 citations, demonstrates how touch can be seamlessly integrated into large-scale multimodal models. By enabling systems to infer tactile properties from visual cues and vice versa, Park’s contributions are foundational for applications in robotic manipulation, prosthetics, and virtual reality. His approach not only reduces the need for exhaustive tactile data collection but also paves the way for more intuitive human-robot interaction. As an early-career researcher, Park’s innovative synthesis of touch with other modalities marks him as a key architect of the next generation of perceptually aware AI.
Research Focus
Key Achievements
Top Papers
- 1