Making Canvas Prints From Artwork Photos
summary
Turns phone-quality listing photos into saleable canvas prints with a production pipeline that enforces opt-in consent and automatic royalties, serves tens of thousands of artists, and launched print‑on‑demand in March 2025. Steps: Clarifi classifier for printability, Claudinary smart cropping, on‑first‑purchase upscale via topaz labs image API, mirror-bleed extension using a Ruby service with lip whips, orientation markers, then send assets to Visttor Print. System works at scale while minimizing artist friction and ethical risk.
AI-written summary · about
chapters (17)
- 0:00Introduction
- 1:30Phone photos to canvas prints
- 5:46Ethical concerns in print on demand
- 8:16Artist concerns addressed
- 10:45Photo to canvas pipeline
- 13:17Smart cropping and upscaling
- 14:24Print production constraints
- 15:53Bleed and orientation
- 16:54Production partner and delivery
- 17:17What upscaling means
- 18:00Traditional interpolation
- 19:06CNN upscaling for fidelity
- 20:34GAN upscaling for realism
- 22:47Diffusion models for refinement
- 24:35Choosing upscaling method
- 25:18Project lessons learned
- 26:23Closing and call to action
related talks
19mKanvas: An NFT marketplace on Tezos
19mAI-Powered Startup Toolkit for a Product Designer
17mCreating an artwork finder with the Cloud Vision API
37mEnhancing Media Experiences in the Age of AI: A Practical Guide
30mMachine learning applications for rock art research
32mSelf-Portraits in JavaScript
27mBeing Smart with Art
25mHow Mixbook leveraged genAI to offer personalized experiences