Youngbin Park

About

Youngbin Park’s research lies at the intersection of robotic manipulation, deep reinforcement learning (RL), and computer vision, with a primary focus on enabling robots to grasp diverse objects in cluttered, real-world environments. His most significant contributions center on accelerating actor-critic deep RL for visual grasping by developing state representation learning methods that preprocess raw visual input, dramatically improving training stability and performance. Notably, his 2021 paper on this topic has garnered 14 citations, while his work on sim-to-real transfer—using pixel-level and feature-level domain adaptation to minimize real-world data requirements—has received 9 citations. Park has also advanced pre-grasping pose prediction by combining deep convolutional neural networks with mixture density networks, and tackled object singulation through nonlinear pushing to clear clutter before grasping. His early work includes map-building and localization using 3D features for service robots, and he has explored semantic segmentation via recurrent convolutional-deconvolutional neural networks that integrate top-down and bottom-up visual processing. Through these efforts, Park has established himself as a key contributor to making deep RL practical for real-world robotic grasping.

Research Focus

Key Achievements

Top Papers

1. 1
   Acceleration of Actor-Critic Deep Reinforcement Learning for Visual Grasping by State Representation Learning Based on a Preprocessed Input Image

   14 citations · 2021
2. 2
   Sim-to-Real Visual Grasping via State Representation Learning Based on Combining Pixel-Level and Feature-Level Domain Adaptation

   9 citations · 2021
3. 3
   Acceleration of Actor-Critic Deep Reinforcement Learning for Visual Grasping in Clutter by State Representation Learning Based on Disentanglement of a Raw Input Image

   7 citations · 2020
4. 4
   Predicting Multiple Pregrasping Poses by Combining Deep Convolutional Neural Networks with Mixture Density Networks

   4 citations · 2016
5. 5
   Object Singulation by Nonlinear Pushing for Robotic Grasping

   3 citations · 2019
6. 6
   Development of an Actor-Critic Deep Reinforcement Learning Platform for Robotic Grasping in Real World

   3 citations · 2020
7. 7
   Map-Building and Localization by Three-Dimensional Local Features for Ubiquitous Service Robot

   3 citations · 2007
8. 8
   Integration of top-down and bottom-up visual processing using a recurrent convolutional–deconvolutional neural network for semantic segmentation

   3 citations · 2019

Key Collaborators

Contact & Links