[
    {
        "role": "user",
        "content": "using%20jaws%20how%20can%20I%20get%20a%20description%20of%20%20images"
    },
    {
        "role": "assistant",
        "content": "Here\u2019s how you can get descriptions of images using **JAWS** screen reader on Windows. I\u2019ll cover both classic methods and the newer AI-powered options that Freedom Scientific added recently.\n\n---\n\n## 1. Alt Text and Basic Graphic Navigation\n\n- When an image is coded with **alt text** (alternative text), JAWS will read that description when the image is focused (e.g., when navigating with graphics commands). ([srt.csb-cde.ca.gov](https://srt.csb-cde.ca.gov/jaws/jaws-lesson-9.html?utm_source=openai))  \n- Use the **G** key to move to the next image (\"graphic\"), and **Shift + G** to move to the previous image. If the image has alt text, it reads it; if not, it might just say \u201cgraphic\u201d or \u201cimage.\u201d ([srt.csb-cde.ca.gov](https://srt.csb-cde.ca.gov/jaws/jaws-lesson-9.html?utm_source=openai))  \n- You can adjust whether JAWS announces *only tagged images (with alt text)* or *all graphics*. To change this, press **INSERT + V** (Quick Settings), look for **Graphics Show - Tagged**, and select *All* if you want every image announced\u2014even those missing alt text. ([www2.freedomscientific.com](https://www2.freedomscientific.com/training/Surfs-Up/difficult_pages.htm?utm_source=openai))  \n\n---\n\n## 2. Image Descriptions with PictureSmart & Generative AI\n\nAs of the **March 2024 update** (build 2402), JAWS supports **PictureSmart AI**, which allows more detailed, on-demand descriptions using AI. ([wcblind.org](https://wcblind.org/2024/04/new-ai-feature-in-jaws-and-fusion-is-worth-a-thousand-words/?utm_source=openai))  \n\nTo use it:\n\n1. Enable the feature:  \n   - Open JAWS, go to the **Early Adopter Program** dialog, check *PictureSmart with generative AI*, and then restart JAWS. ([sastltd.zohodesk.eu](https://sastltd.zohodesk.eu/portal/en/kb/articles/jaws-picturesmart-feature?utm_source=openai))  \n2. Use the layered PictureSmart shortcut:  \n   - Press **INSERT + SPACEBAR**, then **P**, then one of the following letters to describe:\n     - **C** \u2014 the currently focused control  \n     - **W** \u2014 the current window  \n     - **S** \u2014 the visible part of the entire screen  \n     - **F** \u2014 a photo/file currently highlighted in File Explorer ([wcblind.org](https://wcblind.org/2024/04/new-ai-feature-in-jaws-and-fusion-is-worth-a-thousand-words/?utm_source=openai))  \n\nUsing an **uppercase letter** at the end (e.g. **P** instead of **p**) requests a *more detailed description* using AI; may even combine results from multiple engines. ([sastltd.zohodesk.eu](https://sastltd.zohodesk.eu/portal/en/kb/articles/jaws-picturesmart-feature?utm_source=openai))  \n\n---\n\n## 3. Recognizing Text in Images (OCR) and PDF\n\nIf an image is a scan of text (like a scanned document or screenshot), JAWS can use OCR to extract the text and speak it aloud:\n\n- Use **INSERT + SPACEBAR**, then **O**, followed by another letter to choose what area to OCR (e.g. window, screen, etc.). ([support.freedomscientific.com](https://support.freedomscientific.com/Content/Documents/Manuals/JAWS/JAWS-Quick-Start-Guide.pdf?utm_source=openai))  \n- In PDFs, if the document is image-only (no embedded text), the PDF reader may prompt you to tag it or use recognition tools to make it accessible. Otherwise JAWS cannot read what it doesn\u2019t detect as text. ([wwwimages2.adobe.com](https://wwwimages2.adobe.com/content/dam/cc/en/accessibility/pdfs/accessing-pdf-sr.pdf?utm_source=openai))  \n\n---\n\n## 4. Tips & Common Scenarios\n\n- If an image lacks alt text, JAWS won\u2019t have much to read\u2014it may simply announce \u201cgraphic\u201d or \u201cunlabeled graphic.\u201d ([liptrot.org](https://liptrot.org/guides/jaws/images/?utm_source=openai))  \n- If the image is decorative, authors can mark it with empty alt text so that JAWS skips it entirely. ([online.math.uh.edu](https://online.math.uh.edu/jaws-tutorial/?utm_source=openai))  \n- In chat apps like X (formerly Twitter), there are script-specific commands (e.g. `Ctrl + Windows + I`) to trigger image scanning/description. ([hartgen.org](https://www.hartgen.org/XWithJAWS?utm_source=openai))  \n\n---\n\nIf you tell me which version of JAWS you have (or if you're in Early Adopter Program or not), I can give you exact steps for your setup. Want me to look that up for you?"
    }
]