Byung Wan Kim

About

Byung Wan Kim is a researcher specializing in computer vision and deep learning, with a particular focus on semantic segmentation and visual processing architectures. His most notable contribution is the development of a recurrent convolutional–deconvolutional neural network that integrates top-down and bottom-up visual processing, a framework that bridges high-level semantic understanding with low-level spatial detail. This work, published in 2019, has garnered 3 citations and represents a significant step toward more biologically inspired and context-aware segmentation models. Kim’s research addresses a fundamental challenge in computer vision: how to combine global context with fine-grained local features for accurate pixel-level classification. By proposing a recurrent mechanism that iteratively refines segmentation maps, his approach has influenced subsequent work on attention-based and feedback-driven architectures. Though early in its citation trajectory, this paper is recognized for its conceptual novelty and potential applications in autonomous driving, medical imaging, and scene understanding. Kim’s work continues to inspire researchers seeking more robust and interpretable visual recognition systems.

Research Focus

Key Achievements

Top Papers

1. 1
   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