Lmst

Cette semaine sur Oxytude, l'actu des nouvelles technologies et de l'accessibilité.

#Apple #VoiceOver #iOS #macOS #watchOS #visionOS #GoogleGemini #TalkBack #ScanVox #BeMyEyes #BeMyAI #JAWS #ZoomText #Fusion #MicrosoftEdge #Samsung, #Électroménager

https://www.oxytude.org

Five years ago today I released one of my more unusual EP's called Ambiental. It's a strange mix of vocals, sliding noises and long evolving chords. Great for relaxing, meditation, or just to put on in the background and forget about it. Available on all the usual streaming platforms and distributed via #DistroKid. https://distrokid.com/hyperfollow/andrelouis/ambiental
Alt-text courtesy of #BeMyAI

A poster with a central image of a person lying on their side, surrounded by lush green foliage. The person is adorned with flowers, including red roses on their head and a white flower near their hand. A blue butterfly rests on their foot. The text "ALBUM ANNIVERSARY" is repeated multiple times in bold, outlined letters at the top and bottom. The bottom section also includes the text "DISTRIBUTED BY DISTROKID" and "AVAILABLE ON" followed by icons for Spotify, Apple Music, and YouTube. The background has a gradient of light blue to pink.

I asked her to take a picture so I could share the goodness with you all. Image description courtesy of #BeMyAI because I'm no good at explaining stuffs.

Four small cakes are placed on a paper towel. They appear to be mini sponge cakes or muffins, each split in half and filled with a cream or custard. The tops are dusted with powdered sugar. The pastries are arranged in a casual manner, and the lighting suggests they are in a kitchen setting.

TyfloPrzegląd Odcinek nr 265

https://tyflopodcast.net/tyfloprzeglad-odcinek-nr-265/

A fabulous tool for all Mastodonners who post pictures: Be My AI.

The software, available for Windows, Android and iPhone takes a picture or image and provides a very detailed description that you can copy into your Alt text (see examples below).

The interface is extremely simple and the program works for art, nature and graphs.

Adding a good Alt text has never been so easy.

Download it here:
https://www.bemyeyes.com/download

#Alt #AltText #Convenient #LowVision #Tools #Mastodon #BeMyAI

The image shows a building with a striking and creative mural painted on its facade.

The mural is designed to create an optical illusion, making the building appear as if it is melting or distorted. The windows and architectural elements are painted in a wavy, surreal style, giving the impression that the building is bending and warping.

The windows are painted with bright blue, and there are painted green plants on some of the window sills, adding to the illusion.

The building next to it has balconies with real plants, providing a contrast to the painted facade.

The sky is clear and blue, and there is a sign on the right side of the image with the text "AGENZIA VIAGGI".

The image is an infographic titled "Top Countries by Fossil Fuel Consumption 2023." It shows a bar chart comparing the fossil fuel consumption of various countries and regions, measured in exajoules. The chart includes three types of fossil fuels: coal, natural gas, and oil, each represented by different colors.

The countries and regions listed, along with their total consumption in exajoules, are as follows:

- United Kingdom: 5 exajoules
- Germany: 9 exajoules
- Indonesia: 9 exajoules
- Canada: 10 exajoules
- South Korea: 10 exajoules
- Saudi Arabia: 12 exajoules
- Iran: 12 exajoules
- Japan: 15 exajoules
- Russia: 27 exajoules
- India: 35 exajoules
- United States: 76 exajoules
- China: 140 exajoules
- Rest of the world: 146 exajoules

The infographic notes that 1 exajoule is equivalent to 174 million barrels of oil equivalent. It also mentions that China includes Hong Kong SAR. The source of the data is the Energy Institute's 2024 Statistical Review of World Energy, as reported by Business Insider.

The image shows a group of deer standing on a grassy field. There are several deer, with the one in the foreground having large, impressive antlers.

The deer have a mix of brown and white fur, with some having spots on their backs.

In the background, there are trees and a gently sloping hill, suggesting a park or natural reserve setting.

The scene appears peaceful and natural.

I frequently see people asking, even demanding, that AI-generated Alt Text stop. If those folks clearly have usable vision, I block them and move on – life’s too short for preachy whining from people who don’t rely on Alt Text.

But when it’s a blind person saying this, I’m surprised. As a blind person myself, I know that 99% of the images I encounter online lack any description. Now, with a suite of AI tools, I can finally generate accessible, usable image descriptions. Are they perfect? No. Computer vision isn’t sight, but it’s made staggering progress – beyond what I ever expected. And if I need more clarity, I can always consult a human through these same tools.

Blind people slamming AI-generated Alt Text isn’t helping. We need these systems to improve, not disappear. The reality is, expecting every image to have a human-created description is unrealistic. Let’s push for better technology, not bash what’s already making a difference.

#AI #AltText #Blind #BeMyAI #BeMyEyes #AccessAI #ComputerVision #AssistiveTech

Hmm, definitely like that... Being able to ask #Siri to take a #screenshot of something, then sending it over to #BeMyAI for a detailed description. #A11Y #Accessibility #iOSAccessibility #ImageDescriptions

Rachel had a little lamb, little lamb, little lamb! Photo description courtesy of #BeMyAI: The image shows a woman sitting on a chair indoors, smiling with her eyes closed. She is wearing a mustard yellow sweater and light blue jeans. In her arms, she is holding a small lamb with a black face and ears, and a white body with some black patches. The setting appears to be a rustic room with corrugated metal walls. In the background, there is a wooden shelf with some small toy trucks and a large orange pot. A bottle of "Cleenol" is visible on a small table or cabinet behind her.

One cool thing about this morning’s vaccination clusterfuck: I was able to use #BeMyEyes and #BeMyAI to independently browse the soda coolers and get the right pop I wanted. What kind of pop? Let’s just say I got the right one baby! Uh huh! Maybe someone out there is old enough to get the reference.

In the last two weekends, I've used Be My AI in the Be My Eyes app to set the temperature on my thermostat. #BeMyAI #BeMyEyes #accessibiluty

@cachondo I wonder if Be My AI in the Be My Eyes ap could help with this. #BeMyAI #BeMyEyes #accessibility

I decided to snaffle an image from an email simply called 'image.png' to see what #BeMyAI made of it, and got a pretty comprehensive result. Curious to see whether my sighted compatriots agree with the #AltText?

The image shows a detailed interface of a virtual synthesizer, specifically an emulation of the ARP 2600, a classic analog synthesizer. The interface is divided into several sections with numerous knobs, sliders, and switches for various sound parameters.

Key sections include:

- VCO (Voltage Controlled Oscillator) sections labeled VCO 1, VCO 2, and VCO 3, each with controls for frequency, waveform, and modulation.
- VCF (Voltage Controlled Filter) section with controls for filter type, frequency, and resonance.
- VCA (Voltage Controlled Amplifier) section with gain and pan controls.
- Mixer section for combining audio signals.
- Effects section on the right side, listing effects like Modern Chorus, Modern Phaser, and Tape Echo, with adjustable parameters.
- Additional sections for preamplifier, envelope follower, ring modulator, noise generator, and more.

The top part of the interface has a menu bar with options like Main, Effects, Reverb, Mod Matrix, MIDI Map, and Librarian. There are also controls for parameters like frequency, depth, phase, and wet/dry mix.

Comparing AI-Powered Accessibility Tools: PiccyBot, Be My AI, and Aira's Access AI

Introduction
AI-powered tools like PiccyBot, Be My AI, and Aira's Access AI are making significant strides in enhancing accessibility for visually impaired individuals. This article explores and compares the capabilities of these three tools, highlighting their unique features and benefits.

PiccyBot: Comprehensive Image and Video Descriptions
PiccyBot is an AI-driven application that describes images and videos. It provides detailed spoken descriptions, allowing users to understand visual content through audio feedback.
Key Features:
• Photo to Voice: Converts photos into detailed spoken descriptions.
• Voice-Activated Queries: Users can ask questions about the images using their microphone.
• Video Descriptions: Offers descriptions for videos up to 30 seconds long, enhancing accessibility for visual media.

PiccyBot stands out for its ability to describe both static and moving images, providing users with a comprehensive understanding of visual content through interactive and detailed descriptions.

Be My AI: Enhanced Visual Interpretation with GPT-4 Vision
Be My AI is a feature within the Be My Eyes app, utilizing OpenAI’s GPT-4 Vision model to deliver detailed descriptions of images and graphics. This AI tool complements the live assistance provided by sighted volunteers.
Key Features:
• AI-Powered Descriptions: Provides comprehensive descriptions of photos, screenshots, and on-screen content.
• Interactive Chatbot: Allows users to ask follow-up questions for more detailed information.
• Cross-Platform Availability: Accessible on both mobile devices and Windows PCs.

Be My AI excels in offering detailed, immediate visual descriptions that enhance users' understanding of digital content. The interactive chatbot feature allows for deeper engagement and a more nuanced understanding of the described material.

Access AI by Aira: Real-Time and Verified Visual Assistance
Access AI is a new feature within the Aira app, providing AI-driven image descriptions along with the option to verify these descriptions through trained visual interpreters.
Key Features:
• AI Image Chat: Users can take or upload photos to receive detailed image descriptions.
• Human Verification: Visual interpreters can verify and correct AI-generated descriptions, ensuring accuracy and reliability.
• Integration with Aira Services: Combines AI capabilities with Aira’s established visual interpreting service for comprehensive support.

Access AI combines the strengths of AI and human intelligence, offering detailed image descriptions with the added benefit of human verification. This ensures higher accuracy and trust in the AI-generated information, making it a reliable tool for various visual tasks.

Interactivity and Detail:
• PiccyBot provides interactive and detailed descriptions of both images and videos, with the ability to zoom in and ask specific questions.
• Be My AI offers comprehensive and interactive descriptions with an AI chatbot that allows for follow-up questions.
• Access AI provides detailed descriptions verified by human interpreters, combining AI efficiency with human accuracy.

Platform and Availability:
• PiccyBot and Be My AI are available on mobile devices, with Be My AI also accessible on Windows PCs.
• Access AI is integrated within the Aira app, available on both iOS and Android, and offers additional human verification for AI responses.

Subscription and Cost:
• PiccyBot offers both free and premium versions with in-app purchases for enhanced features.
• Be My AI is a free feature within the Be My Eyes app.
• Access AI provides free initial access with options for extended use through Aira’s subscription plans.

Conclusion
PiccyBot, Be My AI, and Access AI each offer unique advantages for visually impaired users. PiccyBot and Be My AI leverage advanced AI to provide detailed and interactive visual descriptions, enhancing digital accessibility. Access AI combines AI with human verification, ensuring accurate and reliable information. By understanding the strengths and limitations of each tool, blind or visually impaired users can choose the best option to meet their specific needs and enhance their independence and quality of life.

#AI #Accessibility #PiccyBot #BeMyAI #AiraAccessAI #VisualImpairment #Blind #TechForGood #AssistiveTechnology #InclusiveTech #AIForAccessibility #TechComparison #AIInnovation #EmpowermentThroughTech #DigitalAccessibility #AIInclusion #TechReviews #AccessibleTechnology

The image depicts an artistic representation of an eye formed by circuitry and digital components, symbolizing artificial intelligence or machine vision. The eye is rendered in striking shades of pink and teal, with circuit board patterns forming the iris and pupil. The title "The Blind AI Blog" appears below, creating an intriguing contrast between vision and blindness. The image conveys themes of technology, perception, and the intersection of human and artificial intelligence in a visually striking and thought-provoking manner.

@cnet #AltText courtesy of #BeMyAI:

The image shows three people standing together indoors, each holding up a badge that says "Media." The person on the left is a woman with long blonde hair, wearing a floral-patterned shirt and a light-colored backpack. The person in the middle is a man with a black headscarf, wearing a gray t-shirt with a graphic of "The Bel-Air Bunch" and holding a camera or some equipment. The person on the right is a woman with bright pink hair, wearing a beige sleeveless dress. The background features a wall with a pattern that looks like a brick wall with small lights embedded in it, and there are other people in the background, some of whom are also wearing badges.

Using a combination of #MPCHC and #BeMyAI I thought I'd take a screenshot of the tennis match and see what image description I got out of it. #AltText response included in post.

Alt-text courtesy of Be My AI.

The image shows a tennis match being played on a grass court. The view is from a high angle, looking down the court. Two players are visible: one near the bottom of the image, closer to the camera, and the other near the net, further away. The player closer to the camera is in the middle of a backhand swing, while the player near the net is moving to the right.

In the background, there are several officials and ball boys/girls standing along the back wall. The scoreboard in the top left corner shows the names "Alcaraz" and "Humbert" with the current score. Alcaraz has 2 sets, 3 games, and 30 points, while Humbert has 1 set, 2 games, and 40 points. The speed of the last serve is displayed as 120 mph.

The bottom part of the image shows the interface of a media player, indicating that the video is currently playing. The media player displays technical details such as frame rate, sync offset, frames drawn and dropped, jitter, buffers, and bitrate. The timestamp at the bottom right corner shows 06:44.

Our neighbors gave our younger daughter a birthday card. I used #BeMyEyes to describe it and it was even able to read the handwritten note! #accessibility #BeMyAI #AI

Most of the time, I can't get the screen portion of #BeMyAI to work because of the empty, secondary app that seems to linger in strange ways. Can sometimes minimize it but I think it still has an odd and adverse affect on the output. Very weird design. That secondary app appears to be responsible for global hotkeys.

haha in other news, don't you just love that flurty-sounding female voice that has been demoed with chatGBT4 O? man it's hilarious! I can't wait for this. I downloaded GBT on my phone and noticed that it's in the share sheet, meaning, I can now have photos described with either #BeMyAI, #SeeingAI or #OpenAI

Having to use #BeMyAI desktop app to work out what some things are, as they can prove interesting in some cases.

I tried an experiment tonight. I put a screenshot of a #Stormshield map into #BeMyAI, and I’d say I got some pretty decent interpretation. #Accessibility #Weather #QuadCities #storms #WeatherRadar
https://share.bemyeyes.com/chat/DFquHwCzKq

#BeMyAI

Client Info