by Simone C Niquille /Technoflesh
Format: HD Video, 3D Animation
Duration: 6 min 50 sec
Audio: Stereo Sound
Produced with the support of the Pax Media Art Award 2020.
In the video work Sorting Song a selection of domestic objects are displayed in perpetual ambiguity. Morphing between a vase and a bowl, a bench and a couch, a chair and a toilet, the fuzzy borders between language, categorisation and objects are explored. “Must a name mean something?” asks Alice in Lewis Carroll’s Through the Looking Glass. The world is constructed through constant negotiation. Where does the bowl end and the vase begin? Perhaps seemingly trivial, these object boundaries are essential to computer vision training datasets that aim to sort objects into neat categories. With whom does the machine negotiate?
Sorting Song features objects from the SceneNet RGB-D indoor training dataset. This dataset by Imperial College London is a large scale repository of 3D meshes, floorplans and objects, compiled to develop computer vision for future domestic robots. The difficulties of sorting objects by their form alone becomes apparent when a wheelchair, toilet and electric chair are all found within the category “chair”. Context matters, yet this data is stripped of it, and training robots that will ultimately cohabit with humans.
As a succession of the video piece Homeschool, Sorting Song picks up where Homeschool left off: at the frustration with the limitation of computed language and the finite world it produces. The work makes visible the protocols and data that shape the digital representation of the world. Taking the format of an educational kids song, the video piece Sorting Song aims to push beyond what appears innocent and naïve, unsettling a world of coded assumptions.
• Tomatoes are fruits and other tales from category-land
A conversation on categorisation with professor of psychology Eleanor Rosch
Transmediale Almanac 2021
• Research, Script & Animation: Simone C Niquille
• Music: Jeff Witscher & Peter Rylander
• Voice Over: Emma Prat & Julian Murray
• Interior & Furniture Assets: SceneNet-RGBD, Dyson Robotics Lab at Imperial College London
→ With a critical nod to Sesame Street's Sorting Song.