Neural network image recognition

Published 2014-04

Products and services using Neural network image recognition are generally available around 2022

Verdict

hitTwo blind graders agree (D7, rule D37). The auditor agent confirmed it (D11).

Grader A: hit
Grader B: hit
agrees with A, without seeing its verdict
Audit: confirmed
in the fixed audit sample (D11)
Published: hit

Grader A: hit

Google Photos (May 2015) searched users' photos by recognised content (animals, food, buildings) using machine learning, free to consumers; cloud vision APIs followed. Available years before 2022.

Test: D37 Horizons: products or services using neural-network image recognition buyable by ordinary customers by 2024 (earlier counts).

Google Photos (Wikipedia) (2015-05-28): Launched May 2015; machine learning recognises and makes searchable people, places and things (animals, food) in photos.

Grader B: hit

Google Photos launched in May 2015 with neural-network content search; Google Cloud Vision API (2016) and Amazon Rekognition (2016) sold image labelling to businesses. Available well before 2022.

Test: Ordinary consumers or businesses could buy/use a neural-network image recognition product by 2024 (earlier counts).

Picture this: a fresh approach to Photos (2015-05-28): Google Photos consumer launch, searchable by image content without tagging.
Introducing Amazon Rekognition (2016-11-30): Deep-learning image recognition sold as a general AWS service.

All captured fields

id: hz-2014-003
category: Neurotechnology and cognitive technologies
kind: technology
target year: 2022
placed year: 2022
milestones: {"scientifically_viable":2014,"mainstream":2021,"financially_viable":2022}
confidence: high
note: Numbered dot (mainstream point) at 2021.0, on the 2021 line (read as 2021). Bar from 2013.8 to 2021.7: scientifically viable ~2014, financially viable ~2022. The bar reaches the inner edge of the grid (~2014), so scientific viability may be at or before the chart start. Sub-area: Neural network computing.