METHODS AND APPARATUSES FOR PRESENTING CHALLENGES TO TELL HUMANS AND COMPUTERS APART

Patent №

US 7,945,952

Granted

2011-05-17

Filed 2005

Owner

GOOGLE, INC.

AI components

3

ml · nlp · speech

Assignment

Recorded

Dataset

AIPD

2023_r1 edition

Application

11173751

Methods and apparatuses to tell humans and machines apart. In one embodiment, automated determination of whether a human or machine is at the other end of a communication channel is based on whether the correct response to a challenge/question is received from the other end of the communication channel. The challenge is provided in a non-image format, such as a text based questions or a sound clip, to prevent abuses of the communication system by computer software robots. The communication system may be primarily for text and/or voice communications, or be such that the rendering image-based challenges can cause significant disruption in normal flow of usage. In one example, the challenge includes deliberately generated random misspelling. In one example, the challenge is context sensitive, factual based, and/or instructive of an operation in a natural language to obtain a result.

Machine learningNatural languageSpeechH04L 63/08G06F 21/316G06F 2221/2103G06F 2221/2105

AI classification

Natural language1.00
Speech1.00
Machine learning0.99
Knowledge representation0.45
AI hardware0.07
Planning0.05
Vision0.05
Evolutionary computation0.01

Ownership

GOOGLE, INC.

assignment · 167580292

Assignors

BEHFOROOZ, REZA

On an employer assignment, the assignors are typically the inventors.

© 2026 NYSGPT2525 LLC