Urgent.News

What's breaking now, across thousands of outlets.

AI

I tried ChatGPT’s new try-it-on shopping function. It worked for daily outfits, but couldn't handle a Halloween costume.

I tried ChatGPT's virtual try-on feature with jewelry, daily outfits, a wedding dress, and a fairy Halloween costume. Here are the function's limits.

I tested OpenAI's virtual try-on feature on jewelry, daily outfits, a wedding dress, and my cat. The tool worked well for accessories and simple clothing, producing realistic results. However, complex outfits and specific styles like wedding dresses proved challenging, often resulting in inaccuracies in garment appearance and body proportions.

The AI aims to serve as a virtual fitting room, allowing users to upload selfies or photos to see how products would look on them. Users can also search for similar items based on celebrity outfits or screenshots of existing items. The new try-on feature is powered by OpenAI's ChatGPT Images 2.5 model, which offers more realistic lighting and textures.

During testing, the AI excelled at finding accessories that matched a user's original selfie, accurately rendering a single mesh top and a full-body outfit, but struggled with a complex wedding dress and even a Halloween costume. The AI recommended items that were either not a good fit or that altered the user's body proportions in unrealistic ways.

Despite some inaccuracies, the virtual try-on proved more helpful for pets than for humans, as it allowed the reporter's cat to try on an adorable Halloween collar.

Written by urgent.news from Business Insider's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.

Read the original at businessinsider.com →

More in AI

Designing a multilingual clinic receptionist: booking, failures and handoff

Originally published as a Fonix.AI practical guide by AntEngage. This republication uses the same guide content. When the front desk cannot pick up An AI receptionist for clinics answers routine…

  • Front desk determines call routing for AI receptionist
  • AI gathers clinic info like hours, services, doctor schedules
  • AI handles bookings, offers callbacks if slots unavailable

QA Isn’t AI Evaluation

An AI agent prepares an internal report. The report has the right format. The figures are accurate. The conclusions sound reasonable. But the agent used a source it wasn’t allowed to access.

  • AI evaluation differs from traditional QA
  • Evaluating AI requires defining right behavior and criteria
  • Human review needed for complex decision judgments

More from Sunday 4 October →