“Deep Learning Algorithms as Potential Solutions to Challenges in Video Art Preservation” presented by Chow and Chan

Symposium:

• ISEA2024: 29th International Symposium on Electronic Art

• Summit on New Media Art Archiving (short papers)

• Deep Learning Algorithms as Potential Solutions to Challenges in Video Art Preservation

• Fourth Summit on New Media Art Archiving

• John Ho Fung Chow
• Sam Hin Chung Chan

Venue(s):

• Brisbane Convention & Exhibition Centre, Brisbane, Australia

• Video artworks created in analog format present unique preservation challenges due to technological and personnel limitations. Current restoration approaches, including re-transferring from master tapes and digital image processing, are imperfect solutions. This paper explores the use of deep-learning algorithms to recover common defects in analog video art. It emphasizes the crucial need for a paradigm shift of concept in media art preservation for a possible way out from the long drastic debate on digital restoration measurement regarding the physical, aesthetical, and historical authenticity of preserving analog media artwork.

Keywords:

• Analog Media
• video art
• Deep Neural Network
• Machine Learning
• media art archiving
• Restoration

Category:

• Short Paper