My first experiences with recognizing images with Seeing AI were not something related to my schoolwork or anything academic— it was actually for recognizing memes. One of my best friends loves to send me text posts, screenshots, and other images with lots of text in them when we are texting or talking on the phone. It can be challenging for me to read text in images due to small font sizes, and using pinch-to-zoom with horizontal scrolling can get tedious as well, so using AI-powered image recognition tools like Seeing AI can make it easier for me to figure out what is in an image quickly and read text in a format that works for me without missing a beat in the conversation. Here are my tips for recognizing images with Seeing AI, and how I use the image recognition tool specifically in multiple contexts as someone with low vision.
Background: What is Seeing AI?
Microsoft Seeing AI is a free visual assistance app powered by artificial intelligence and machine learning that provides information about objects and text for people with vision loss (inclusive of blind/low vision). With Seeing AI, users can use their device camera to take a photo or point it at something in their environment to get a description, or upload something from their camera roll to get a visual description within seconds. The Seeing AI interface is self-voicing and uses large, bold text by default, so users do not need to enable VoiceOver or other features to get access to the different accessibility settings.
Seeing AI is currently available for iOS and Android and requires iOS 12 or later/Android 6.0 or later. Some channels (features) require internet access to work properly, but there is no requirement to create an account. All users must be 13 years old or older.
Related links
- Seeing AI on the App Store
- Seeing AI – Apps on Google Play
- Microsoft Seeing AI And Low Vision
- All About Visual Assistance Apps For Visually Impaired
How to recognize images with Seeing AI/Upload gallery images to Seeing AI
There are several options for recognizing images with Seeing AI from the device gallery, which can include uploading images directly into the application and using the share sheet/share menu within the gallery. Follow these instructions to upload images to Seeing AI to use the different image recognition features.
Upload image from phone gallery to Seeing AI: Android
- Open image in Photos or other stored location
- Select the Share button
- From the list of options, select Seeing AI
- A loading jingle will play indicating the content is being recognized/uploaded
Upload image from gallery to Seeing AI: iOS
- Open image in Photos or other stored location
- Select the Share sheet icon
- From the list of options, select Recognize With Seeing AI. This may need to be added manually to the share sheet when the app is first installed, but should show up automatically after that
- A loading jingle will play indicating the content is being recognized/uploaded
Upload image within Seeing AI: iOS only
- Open Seeing AI
- Select the Menu icon, which has three horizontal lines
- Select Browse Photos
- Choose desired image from Photos
- A loading jingle will play indicating the content is being recognized/uploaded
Identify embedded image with Seeing AI
- Long-press on an image in a web browser or other image that can be saved to the device
- Select Share/Share Image
- From the list, select Seeing AI or Recognize With Seeing AI
- A loading jingle will play indicating the content is being recognized/uploaded
Can Seeing AI identify content other than images?
Yes, Seeing AI supports uploading files that aren’t images such as PDFs and short videos, though the video recognition feature is not unlimited and has limited support for different file types (to be honest, I haven’t been able to get it to work for any videos yet as of January 2025). Users can upload content with the share button/share sheet to Seeing AI to recognize text and other details.
Can I identify images without saving them to the camera roll?
Yes, the Seeing AI app is a smart camera that lets users take a photo of something in their environment and listen to/read a description, without saving the photo to their camera roll, unless the user selects the option to save a photo to their device.
Related links
- How I Use My Phone As Assistive Technology In Class
- How I Read Research Sources With Assistive Technology
- A to Z of Assistive Technology for Reading Digital Text
- Texting and Low Vision
Seeing AI image recognition interface
When an image is uploaded to Seeing AI, users are taken to a screen where their image is in the center, and there are three options for interacting with the content: Share, Explore Photo, and More Info.
Share with Seeing AI Image Recognition
Selecting the Share button on a processed image will either share the image with another application or share a copy of the text-based description. On Android, the text description is shared as a .txt file to another application, while on iOS users can copy/paste the text-based description by selecting the desired text or select Share Text from the More Info screen.
Exploring photos with Seeing AI
Want to explore individual components of an image, like chat bubble text or individual objects? Using the Explore Photo button allows users to explore an image by swiping/gestures, identifying text and image descriptions for different items on a page. A colored outline appears around areas of interest as the text or descriptions are spoken out loud, with a jingle/auditory feedback playing as a user selects an area. This is really helpful for exploring screenshots of text like menus or social media graphics.
More Info
By default, a text-based description of the image is displayed in the system font size below the image. Users can scroll through the description or select More Info to get an even more detailed description of what is in an image, including more specific descriptions of items or information on background objects. For example, I uploaded a photo of my dinner to Seeing AI, and the initial description said “plate of food on a table.” With More Info, Seeing AI gave me the following description:
“The image depicts a close-up view of a serving of enchilada casserole placed in a shallow, round dish. The dish has a decorative pattern around its rim. The casserole is composed of several layers, with visible ingredients including black beans, slices of yellow squash, and tortillas. The top layer is covered with a red sauce and sprinkled with shredded cheese. The casserole is positioned centrally in the dish, with a portion of the dish’s interior visible around it. In the background, there are blurred elements, including a tablecloth with a floral pattern, a cork coaster, and a purple container.”
Exiting Seeing AI
Once finished with the image recognition feature, users can exit Seeing AI by selecting the Close button or using the back/home button on their device.
Related links
- How To Write Alt Text and Image Descriptions for the Visually Impaired
- How To Access Images Without Alt Text
- How To Write Alt Text And Image Descriptions For Food
- Options For Writing Extended Image Descriptions On Social Media
Recognizing text with Seeing AI
For images with text, Seeing AI will write out all visible text in the image verbatim, though some text formatting and spacing may be ignored. The text will be displayed with Dynamic Text settings so that users can easily read information with large print or with a screen reader. I’ve found that Seeing AI typically ignores blurred out or irrelevant background text, such as a stop sign.
Examples of images that I use text recognition with include:
- Screenshots of text posts from social media (tweets, Tumblr posts, etc.), especially if they have decorative font
- Images that include lots of text or notepad screenshots
- Screenshots of text conversations
- Pictures with captions
- Diagrams with labels
- Text written in another language— Seeing AI supports 16+ languages for descriptions, but does not provide translations
This is especially helpful for reading low resolution images or images with poor contrast text, and it can also recognize handwriting.
Related links
- How To Create Accessible Diagrams For Low Vision
- Language Learning Tips And Resources For Low Vision
- High Contrast and Low Vision
- Reading Handwriting With Assistive Technology
Scene descriptions
What’s going on in that photo? Seeing AI can provide scene descriptions that describe relevant parts of an image. This is a wonderful way to generate alt text and image descriptions, though copy-and-paste capabilities are only available in the iOS application. That said, the scene description feature still has a ton of other uses and is especially great with photos.
Examples of images I use scene description with include:
- Photos or pictures of a science lab
- Photos that my friends send me where I’m not entirely sure what I am looking at— this has saved me from having to play “20 questions” about what is in a picture before being able to comment on it!
- Memes or other images with a mix of art styles or text content included
- Taking a photo of my desk or a table to locate items I haven’t been able to find— sometimes it’s easier to look at a screen
As shown with the enchilada casserole example, the More Info feature can generate very detailed scene descriptions and descriptions of what is in an image. While it isn’t 100% perfect (since my enchiladas didn’t actually have shredded cheese), it can be helpful for getting context or clarification about what is in an image instead of just relying on my eyes alone.
Related links
- How To Write Video Descriptions For Animal Videos
- How To Write Alt Text For Amateur Art
- How To Write Image Descriptions For Buildings and Architecture
- How To Write Alt Text For Digital Comics
- Lemon On A Pear: Accessible Memes For Low Vision
- How To Describe Science Experiments For Visually Impaired Students
Descriptions of people
For photos with people looking at the camera, Seeing AI can provide descriptions of what each person in the photo looks like, including the following information:
- Approximate age
- Gender
- Hair color
- Any identifying features, i.e glasses
- Their facial expression, such as happy or neutral
In my experience, Seeing AI does not provide descriptions of clothing, unless More Info is selected. In addition, faces that are imported into the app’s facial recognition feature aren’t identified when using the image recognition feature in the photo gallery— so if I upload a picture of my friend K, Seeing AI won’t tell me that the photo has K in it, it will just describe what they look like.
Related links
- How To Create Helpful Visual Descriptions For Visually Impaired Audiences
- My Thoughts On Visual Descriptions With Low Vision
- How Tinted Glasses Help My Light Sensitivity
More ways that I use image recognition with Seeing AI and low vision
- Seeing AI’s image recognition tools are invaluable when I encounter images without alt text— learn more in How To Access Images Without Alt Text
- Another way Seeing AI has helped me a lot is by identifying emoji in a screenshot my friend sent me. I write more about emoji in Emoji Accessibility and Low Vision
- One time, Seeing AI alerted me to the fact that a sign with my hotel name was in the background of a selfie, which would have leaked where I was staying if I posted the image on social media. Good thing I checked— learn more about taking selfies in Taking Selfies With Low Vision
- Seeing AI can’t look up images online, but Google Lens can. I use both apps in different contexts, and talk more about Lens in How I Use Google Lens With Low Vision
Published April 21, 2020. Updated January 2025