#image-analysis

[ follow ]
#ai #openai #ai-model-updates #competitive-rivalry #military-ai-policy #large-language-models #color-thief
Artificial intelligence
fromArs Technica
1 week ago

OpenAI introduces GPT-5.4 with more knowledge-work capability

OpenAI released GPT-5.4 with improved image analysis up to 10.24 million pixels and 18% fewer factual errors, competing against Anthropic's recent user gains from military policy disputes.
Web design
fromRaymondcamden
1 week ago

Dyanimically Adjusting Image Text for Contrast

Color Thief library now includes TypeScript definitions and new contrast detection features, including a textColor property that automatically suggests optimal text color for readability over images.
Privacy technologies
fromPrivacy International
2 weeks ago

Nowhere to Hide? Privacy Risks and Policy Implications of AI Geolocation

Vision-Language Models can accurately determine photo locations without GPS data, creating serious privacy and human rights risks including surveillance, doxxing, and discriminatory policing.
Photography
fromTechRepublic
1 month ago

6 Fun Questions That Make Google Photos Feel Like Magic

Google Photos' new Ask button lets users query photo libraries in natural language to identify locations, describe images, transcribe text, and request edits.
Science
fromMail Online
5 months ago

Shroud of Turin mystery deepens as scientist spots hidden clue

A dentist identified faint lower front teeth on the Shroud of Turin, proposing they support a resurrection-energy imprint while some experts remain skeptical.
fromHackernoon
1 year ago

Why Historical Documents Are Hard to Digitize | HackerNoon

Uneven illumination in optical imaging leads to the diminishing of incident light, making document image analysis, especially character recognition using OCR, challenging. Factors like background objects and light scattering contribute to this problem.
Digital life
#ai
more#ai
[ Load more ]