Shinseungback Kimyonghun's 'Flowers' Challenges Machine Vision
The Korean artist duo Shinseungback Kimyonghun created the series 'Flowers' (2016-17) to test the limits of Google's Cloud Vision algorithm. They collected flower images that the AI recognized with high confidence, then digitally distorted them into pastel abstractions. While humans saw only abstract forms, the algorithm persisted in labeling them as flowers. The work, featured in an article by Dominique Moulon on artpress.com (March 16, 2021), critiques the autonomy of machine vision and its divergence from human perception, raising questions about surveillance and opacity in public spaces.
Key facts
- Shinseungback Kimyonghun is a Korean artist duo comprising Shin Seung Back and Kim Yong Hun.
- The series 'Flowers' was created in 2016-17.
- The work uses Google's Cloud Vision API to analyze images.
- The artists collected flower images that the AI recognized with certainty.
- They then distorted the images into abstract, pastel-colored forms.
- Humans perceived the distorted images as abstract, but the AI still classified them as flowers.
- The article is part of a series on art and technology, following pieces on Anna Ridler and Sabrina Ratté.
- The work questions the autonomy of machine vision and its implications for surveillance.
- The article was published on artpress.com on March 16, 2021.
- The author is Dominique Moulon.
Entities
Artists
- Shinseungback Kimyonghun
- Shin Seung Back
- Kim Yong Hun
- Anna Ridler
- Sabrina Ratté
- René Magritte
- Dominique Moulon
Institutions
- artpress.com
Sources
- artpress —