Special Session 1

Depth Estimation in Vision Systems for Real-time Industrial Applications

Organizer: Assoc. Prof. Muhammad Tariq Mahmood, Korea University of Technology and Education (KOREATECH), Korea

This session aims to explore advances in depth estimation techniques tailored for high-speed, high-accuracy industrial vision systems. As smart manufacturing and Industry 4.0 evolve, depth estimation plays a critical role in enabling automation, quality control, robotic navigation, and 3D inspection. This session welcomes contributions that address the algorithmic, hardware, and system-level challenges in real-time depth estimation under industrial constraints such as variable lighting, occlusions, high throughput, and limited computational resources.

Topics of Interest (but not limited to):

Depth from Stereo, Focus, Defocus, or Structure-from-Motion

Deep learning-based monocular or multi-view depth estimation

Real-time 3D reconstruction and SLAM in factory settings

Depth estimation for robotic pick-and-place, navigation, or inspection

Lightweight and embedded depth estimation models

Hybrid systems: LiDAR + vision, ToF + AI fusion

Benchmark datasets and evaluation protocols for industrial tasks

Hardware acceleration (GPU, FPGA, edge AI for depth processing)

Industrial deployment case studies and system integration

 Submission Link: https://www.zmeeting.org/Submit/paper/track_id/292/short_url/APCT2026.html