Welcome to the 4th edition of the Geometry meets Deep Learning (GMDL) workshop held with ICCV 2019. This series of workshops was initiated at ECCV 2016.

The goal of this workshop is to encourage the interplay between geometric vision and deep learning. Deep learning has emerged as a common approach to learning data-driven representations. While deep learning approaches have obtained remarkable performance improvements in most 2D vision problems, such as image classification and object detection, they cannot be directly applied to geometric vision problems due to the fundamental differences between 2D and 3D vision problems, such as the non-Euclidean nature of geometric objects, higher dimensionality, and the lack of large-scale annotated 3D datasets. Designing geometric components or constraints to improve the performances of deep neural networks is also a promising direction worth further exploration. The workshop aims to bring together experts from 3D vision, graphics and deep learning areas to summarize recent advances, exchange ideas, and inspire new directions.

The workshop will consist of invited talks, spotlight presentations, and a poster session. We are soliciting original contributions that deploy deep learning, 3D geometry, and optimization techniques to solve geometric vision problems including, but not limited to:

• 3D reconstruction of objects, humans and scenes
• 6DoF pose estimation
• Correspondence problems
• 3D shape analysis
• 3D scene understanding
• Visual localization and SLAM
• Geometric deep learning
• Data mining and signal processing on graphs
• Deep learning on manifolds and non-Euclidean domains

Please refer to the submission page for the submission guidelines.

For any inquiries regarding the workshop, please contact Kosta Derpanis at kosta@ryerson.ca