I propose a new approach to understanding how human vision interprets 3D shape in realistic pictures, integrating ideas from human vision science, art history, and computational photography. This theory suggests new ways to think about how pictures work, and how we can make them.

Scientific Papers

• A. Hertzmann. Toward a Theory of Perspective Perception in Pictures. Journal of Vision. April 2024, 24(4). [Paper]
  The main paper
  
  • Condensed version
    I recommend reading this much shorter-version before the main paper. It focuses on the main proposed hypotheses for perspective, and omits most of the literature survey.

• A. Hertzmann. The Choices Hidden in Photography. Journal of Vision. October 2022, 22(10). [Paper]
  This short paper argues that photographs are not "objective" depictions of reality, because of tone, color, and perspective.

• A. Hertzmann. Comparing Perspective in Drawings, Photographs, and Perception. Art & Perception. 2025. [Preprint] [Paper]
  Observations about how my drawings relate to photographs and visual experience, and how similar trends appear in art history.

Related photo manipulation papers

• K. Zhang, J.-B. Huang, J. Echevarria, S. DiVerdi, A. Hertzmann. MaDCoW: Marginal Distortion Correction for Wide-Angle Photography with Arbitrary Objects. CVPR 2025. To Appear.
• S. J. Liu, M. Agrawala, S. DiVerdi, A. Hertzmann. ZoomShop: Depth-Aware Editing of Photographic Composition Computer Graphics Forum, 41(2), 2022. [Paper (Low-res version)]

Blog posts

You can read the following blog posts in order to get the main ideas of the paper:

• The Illusion of Awareness
• Perspective Distortions: Why Normal Cameras Make Faces Look Weird
• Distortion and multiperspective in art and photography
• How we see shape in pictures: perspective: Summary of the whole theory

Other related blog posts:

• Photography is not "objective"
• Visual Illusions Explainable by the Limitations of Peripheral Vision
• How does perspective work, old post from before this work that discusses "natural perspective"

Related Papers

Recommended readings. These are the scientific and technical papers that I found most inspiring and informative for this work. I recommend these as starting points for a deeper dive into the background of the paper's topics.

Talk

Keynote at ICCP 2023:

(this is an older version that does not include fragmentary vision concepts at the end).

"Pearblossom Highway., 11 - 18th April 1986, #2" by David Hockney