Enlarge / Whatever you do, don't ask the AI how many horizontal lines are in this image. (credit: Getty Images) In the last couple of years, we've seen amazing advancements in AI systems when it comes to recognizing and analyzing the contents of complicated images. But a new paper highlights how many state-of-the-art "vision learning Models" (VLMs) often fail at simple, low-level visual analysis tasks that are trivially easy for a human. In the provocatively titled pre-print paper "Vision language models are blind" (which has a PDF version that includes a dark sunglasses emoji in the title), researchers from Auburn University and the University of Alberta create eight simple visual acuity tests with objectively correct answers.

Topics:  enlarge    whatever   getty images    models vlms    vision   pdf   auburn university   alberta   ais   rahmanzadehgervi   bolton   taesiri   nguyen   crucially   comments   ai   vlms   university   basic   tests   visual   shapes   lines   simple   state-of-the-art   credit   researchers   identifying   solve   paper   
BING NEWS:
  • Apple ships truly open-source AI models
    The Apple release is just the latest in a string of such releases to have emerged since the company intensified its focus on AI research. The company has now published dozens of models, including most ...
    07/31/2024 - 4:00 am | View Link
  • The tech industry wants to create an AI utopia. Its leaders think Universal Basic Income is the answer.
    Figures including Sam Altman think a form of universal basic income, or UBI, could be a solution if artificial intelligence starts destroying jobs.
    07/30/2024 - 12:44 am | View Link
  • Zuckerberg touts Meta’s latest video vision AI with Nvidia CEO Jensen Huang
    Segment Anything 2 (SA2) will be open and free to use. There's no word of a hosted version, but there is a free demo.
    07/29/2024 - 11:21 pm | View Link
  • Meta releases its biggest ‘open’ AI model yet
    Meta's newest 'open' AI model release is its biggest yet. The company claims the model, Llama 3.1 405B, is competitive with the best commercial releases.
    07/23/2024 - 5:08 am | View Link
  • Beijing appears to be demanding that Chinese AI models have 'socialist values'
    Beijing's aggressive ideological requirements regarding its homegrown LLMs could hold China's AI sector back, one expert said.
    07/22/2024 - 1:33 am | View Link
  • More

 

Welcome to Wopular!

Welcome to Wopular

Wopular is an online newspaper rack, giving you a summary view of the top headlines from the top news sites.

Senh Duong (Founder)
Wopular, MWB, RottenTomatoes

Subscribe to Wopular's RSS Fan Wopular on Facebook Follow Wopular on Twitter Follow Wopular on Google Plus

MoviesWithButter : Our Sister Site

More Business News