Similar Stories to Can You Do Better Than Top-level Ai Models On These Basic Vision Tests? on Bing News

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:
  • Startup Harrison.ai launches radiology-specific language model as a building block for healthcare AI
    Australia-based startup Harrison.ai spent the past four years building out AI-powered medical diagnostic software and services. | Australian health tech company Harrison.ai has unveiled what it ...
    09/4/2024 - 2:00 am | View Link
  • U.S. intelligence agency to evaluate trustworthiness of AI models
    Based in Springfield, Virginia, NGA collects, analyzes and distributes geospatial intelligence derived from satellite and aerial imagery to support national security, military operations and disaster ...
    09/3/2024 - 9:58 am | View Link
  • Microsoft releases powerful new Phi-3.5 models, beating Google, OpenAI and more
    Microsoft’s release of the Phi-3.5 series represents a significant step forward in the development of multilingual and multimodal AI.
    08/20/2024 - 5:41 pm | 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