Preface

Shi-Min Hu, Connelly Barnes, Changhe TuJCST 2021 · DOI

This special section of the Journal of Computer Science and Technology (JCST) features a curated selection of the most significant research presented at the 9th Computational Visual Media (CVM 2021) conference. Held at Shandong University, the conference serves as a premier cross-disciplinary forum, bridging the gap between computer graphics, computer vision, and geometric computing.

The Convergence of Visual Media

The field of "Computational Visual Media" addresses the historical fragmentation between researchers focused on understanding images (computer vision) and those focused on generating them (computer graphics). The CVM series promotes a unified computational approach, recognizing that the boundaries between these disciplines are increasingly blurred by the integration of machine learning with traditional geometric and rendering pipelines.

Technical Selection and Rigor

The selection process for this special section was exceptionally rigorous. The CVM 2021 technical program committee, comprised of 105 international experts and 37 additional reviewers, evaluated 179 submissions. Of these, 41 papers (approximately 23%) were accepted for oral presentation. From this elite group, six "outstanding" papers—representing roughly the top 3% of all submissions—were distilled for publication in this JCST special section.

Research Highlights

The featured papers provide a snapshot of current trends in the field, moving toward high-fidelity synthesis and spatial intelligence:

  • 3D Reconstruction and Spatial Intelligence: Several works explore multi-view stereo (MVS) for digital twin creation, indoor visual localization, and 3D object tracking. These contributions highlight a shift toward embodied AI and robotics applications.
  • Synthesis and Micro-Detail: Research into cloth wrinkle synthesis demonstrates the field's progression toward physically plausible and visually convincing micro-details in virtual environments.
  • Recognition and Semantic Analysis: Specialized detection tasks, including age/gender estimation and scene text detection, showcase the continued evolution of automated image understanding.

Editorial Leadership

The curation of this special section was led by prominent figures in the computational visual media community.

Guest Editor Shi-Min Hu (Tsinghua University)

The guest editors include Shi-Min Hu (Tsinghua University), Connelly Barnes (Adobe Research), and Chang-He Tu (Shandong University). Their leadership ensures that this collection represents the state-of-the-art in the amalgamation of vision, graphics, and geometry.

Summary updated

BibTeX
@article{hu2021preface,
  author = {Shi-Min Hu and Connelly Barnes and Changhe Tu},
  title = {{Preface}},
  journal = {J. Comput. Sci. Technol.},
  year = {2021},
  doi = {10.1007/s11390-021-0003-2},
}
Figure