Applause launches new approach to training and testing for biometrics and AI

Technology testing company Applause has announced the launch of a new solution for training and testing artificial intelligence algorithms with data collected from the company’s global community of testers to provide the widest possible input range. Applause says the solution is scalable, and its results are tested for across every possible device, location and circumstance to identify issues and provide actionable feedback.

The company claims this will help its customers identify quality issues or bias earlier in the development of their AI systems so they can correct them.

“Our customers have been needing additional support from us in the area of data collection to support their AI developments, train their system, and then test the functionality,” Applause VP of Product Kristin Simonini told AI News. “That latter part being more in-line with what they traditionally expect from us.”

Simonini tells AI News that she has recently read a study in which voice recognition accuracy for white males was over 90 percent, but for African-American women, it was closer to 30 percent.

AI News reports that most of Applause’ experience is with companies in the voice space, but it is increasingly focused on gathering and labeling images for computer vision applications.

The five types of AI applications the new Applause solution is designed for include speech recognition, optimized character recognition (OCR), image recognition, biometrics like fingerprints and facial recognition, and chatbots.

“Users want AI to be more natural, more human. Applause’s crowdsourced approach delivers what AI has been missing: a diverse and large collection of real human interactions prior to release,” comments Simonini in the announcement. “Not only will this improve AI experiences for consumers everywhere, the breadth of the community also has the potential to mitigate bias concerns and make AI more representative of the real world.”

Simonini will deliver a keynote on “Why the Human Element Remains Essential in Applied AI” at the AI Expo North America on November 13.

Various tools and datasets have been developed to help reduce AI bias, and NIST is expected to soon release a report specifically on differences in facial recognition performance between demographics.

Article Topics

artificial intelligence  |  biometric dataset  |  biometrics  |  training