Imageability is a measure of how easily a physical object, word or environment will evoke a clear mental image in the mind of any person observing it. It is used in architecture and city planning, in psycholinguistics, and in automated computer vision research. In automated image recognition, training models to connect images with concepts that have low imageability can lead to biased and harmful results.

History and components

Kevin A. Lynch first introduced the term, "imageability" in his 1960 book, '' The Image of the City''. In the book, Lynch argues cities contain a key set of physical elements that people use to understand the environment, orient themselves inside of it, and assign it meaning. Lynch argues the five key elements that impact the imageability of a city are Paths, Edges, Districts, Nodes, and Landmarks. * Paths: channels in which people travel. ''Examples:

street A street is a public thoroughfare in a city, town or village, typically lined with Building, buildings on one or both sides. Streets often include pavements (sidewalks), pedestrian crossings, and sometimes amenities like Street light, streetligh ...

sidewalk A sidewalk (North American English), pavement (British English, South African English), or footpath (Hiberno-English, Irish English, Indian English, Australian English, New Zealand English) is a path along the side of a road. Usually constr ...

s, trails,

canal Canals or artificial waterways are waterways or engineered channels built for drainage management (e.g. flood control and irrigation) or for conveyancing water transport vehicles (e.g. water taxi). They carry free, calm surface ...

railroad Rail transport (also known as train transport) is a means of transport using wheeled vehicles running in railway track, tracks, which usually consist of two parallel steel railway track, rails. Rail transport is one of the two primary means of ...

s.'' * Edges: objects that form boundaries around space. ''Examples:

wall A wall is a structure and a surface that defines an area; carries a load; provides security, shelter, or soundproofing; or serves a decorative purpose. There are various types of walls, including border barriers between countries, brick wal ...

s, buildings,

shore A coast (coastline, shoreline, seashore) is the land next to the sea or the line that forms the boundary between the land and the ocean or a lake. Coasts are influenced by the topography of the surrounding landscape and by aquatic erosion, su ...

line, curbstone, streets, and

overpass An overpass, called an overbridge or flyover (for a road only) in the United Kingdom and some other Commonwealth countries, is a bridge, road, railway or similar structure that is over another road or railway. An ''overpass'' and '' underpa ...

es.'' *

District A district is a type of administrative division that in some countries is managed by the local government. Across the world, areas known as "districts" vary greatly in size, spanning regions or county, counties, several municipality, municip ...

s: medium to large areas people can enter into and out of that have a common set of identifiable characteristics. * Nodes: large areas people can enter, that serve as the foci of the city, neighborhood, district, etc. *

Landmark A landmark is a recognizable natural or artificial feature used for navigation, a feature that stands out from its near environment and is often visible from long distances. In modern-day use, the term can also be applied to smaller structures ...

s: memorable points of reference people cannot enter into. ''Examples: signs, mountains and public art.'' In 1914, half a century before ''The Image of the City'' was published, Paul Stern discussed a concept similar to imageability in the context of art. Stern, in Susan Langer's ''Reflections on Art,'' names the attribute that describes how vividly and intensely an artistic object could be experienced ''apparency.''

In computer vision

Automated image recognition was developed by using machine learning to find patterns in large, annotated datasets of photographs, like

ImageNet The ImageNet project is a large visual database designed for use in Outline of object recognition, visual object recognition software research. More than 14 million images have been hand-annotated by the project to indicate what objects are pictur ...

. Images in ImageNet are labelled using concepts in

WordNet WordNet is a lexical database of semantic relations between words that links words into semantic relations including synonyms, hyponyms, and meronyms. The synonyms are grouped into ''synsets'' with short definitions and usage examples. It can thu ...

. Concepts that are easily expressed verbally, like "early", are seen as less "imageable" than nouns referring to physical objects like "leaf". Training AI models to associate concepts with low imageability with specific images can lead to problematic bias in image recognition algorithms. This has particularly been critiqued as it relates to the "person" category of WordNet and therefore also ImageNet. Trevor Pagan and Kate Crawford demonstrated in their essay "Excavating AI" and their art project ImageNet Roulette how this leads to photos of ordinary people being labelled by AI systems as "terrorists" or "sex offenders".{{Cite journal , last1=Crawford , first1=Kate , author-link=Kate Crawford , last2=Trevor , first2=Pagan , author-link2=Trevor Paglen , date=2019 , title=Excavating AI: The Politics of Images in Machine Learning Datasets , url=https://www.excavating.ai , journal=The AI Now Institute Images in datasets are often labelled as having a certain level of imageability. As described by Kaiyu Yang,

Fei-Fei Li Fei-Fei Li (; born in Beijing, China, July 3, 1976) is a Chinese-American computer scientist known for her pioneering work in artificial intelligence (AI), particularly in computer vision. She is best known for establishing ImageNet, the dat ...

and co-authors, this is often done following criteria from

Allan Paivio Allan Urho Paivio (March 29, 1925 – June 19, 2016) was a professor of psychology at the University of Western Ontario and former bodybuilder. He earned his Ph.D. from McGill University in 1959 and taught at the University of Western Ontario fr ...

and collaborators' 1968 psycholinguistic study of nouns. Yang el.al. write that dataset annotators tasked with labelling imageability "see a list of words and rate each word on a 1-7 scale from 'low imagery' to 'high imagery'. To avoid biased or harmful image recognition and image generation, Yang et.al. recommend not training vision recognition models on concepts with low imageability, especially when the concepts are offensive (such as sexual or racial slurs) or sensitive (their examples for this category include "orphan", "separatist", "Anglo-Saxon" and "crossover voter"). Even "safe" concepts with low imageability, like "great-niece" or "vegetarian" can lead to misleading results and should be avoided.

References

Psychogeography Environmental psychology Knowledge representation

History and components

In computer vision

See also

Further reading

References