Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Make sure we have alt text for all images #75

Open
davanstrien opened this issue May 5, 2022 · 5 comments
Open

Make sure we have alt text for all images #75

davanstrien opened this issue May 5, 2022 · 5 comments
Assignees
Labels
good first issue Good for newcomers

Comments

@davanstrien
Copy link
Member

No description provided.

@chennesy chennesy added the good first issue Good for newcomers label Jan 31, 2024
@RuthBurns
Copy link
Contributor

RuthBurns commented Jan 31, 2024

List of current alt text:

Welcome
https://carpentries-incubator.github.io/machine-learning-librarians-archivists/01-welcome/index.html
No images

Artificial Intelligence (AI) and Machine Learning (ML) in a nutshell

  1. Image: Flow chart "Input, interpret, reason, decide, response, learn," Current Alt text? = Yes "alt="AI Process"

Machine Learning Modelling Concepts https://carpentries-incubator.github.io/machine-learning-librarians-archivists/03-AI-ML-concepts/index.html

  1. Image: Screenshot of table with date highlighted and pop out boxes of how the data drills down.
    img src="../fig/ep-02-date_feature_engineering.png" alt="Date representations"

What is machine learning good at? https://carpentries-incubator.github.io/machine-learning-librarians-archivists/04-what-is-ml-good-at/index.html

  1. Image with robot, and "input, feature maps, output"
    img src="../fig/ep-03-cnn-diagram.png" alt="Wikipedia CNN Diagram"

  2. Image Classification- An image comparing image classication with object detection
    img src="../fig/ep-03-titanic_newspaper_cv.png" alt="Titanic Newspaper CV"

  3. Activity: Photograph of an archive with drawers open showing many dead colourful birds.
    img src="../fig/ep-03-bird-collection.jpg" alt="USNM Bird Collection"

  4. PixPlot image- a screenshot of image fields in the Meserve-Kunhardt Collection
    img src="../fig/ep-03-pixplot-github.png" alt="PixPlot screenshot from GitHub"

Understanding and managing bias https://carpentries-incubator.github.io/machine-learning-librarians-archivists/05-managing-data-bias/index.html

7)- Set of four wedding photos
img src="../fig/ep-04-Google_Inclusive_Activity_1.png" alt="Four separate images depicting wedding ceremonies"
7b) The same image as 7 but with descriptive labelling below.

Applying Machine Learning https://carpentries-incubator.github.io/machine-learning-librarians-archivists/06-applying-machine-learning/index.html

  1. Notepad scribble of processing pages to machine readable text
    img src="../fig/05-business-use-ocr.png" alt="Illustration of a business problem"

  2. Notepad scribble of text to catalogue process
    img src="../fig/05-predictions-to-actions.png" alt="Two possible ways of using Machine Learning Predictions"

  3. Notepad scribble of Digitized newspaper processes
    img src="../fig/05-model-choices.png" alt="Illustration of possible model choices"

  4. Notepad scribble of photograph to preservation process
    img src="../fig/05-using-model.png" alt="A simplified GLAM workflow"

The Machine Learning Ecosystem https://carpentries-incubator.github.io/machine-learning-librarians-archivists/07-ecosystem/index.html
No images

@PhilReedData
Copy link
Contributor

Have we checked that all these images' licenses are included too? (CC BY 4.0 or similar)

@PhilReedData
Copy link
Contributor

Define what makes the alt text fit for purpose. I've checked the Carpentries guidance.

We should check that the surrounding text (before or after) the image is aligned to. For example, describe the trays of birds in episode 5.

Flow of AI in episode1 could expand to say more.

@RuthBurns
Copy link
Contributor

RuthBurns commented Jan 31, 2024

Artificial Intelligence (AI) and Machine Learning (ML) in a nutshell https://carpentries-incubator.github.io/machine-learning-librarians-archivists/02-AI-in-a-Nutshell/index.html

Image: Flow chart "Input, interpret, reason, decide, response, learn," Current Alt text? = Yes "alt="AI Process"
Proposed alt text: A flowchart demonstrating AI Process

Machine Learning Modelling Concepts https://carpentries-incubator.github.io/machine-learning-librarians-archivists/03-AI-ML-concepts/index.html

img src="../fig/ep-02-date_feature_engineering.png" alt="Date representations"
Proposed alt text: Table showing the data represented within a simple date, such as day of the week.

What is machine learning good at? https://carpentries-incubator.github.io/machine-learning-librarians-archivists/04-what-is-ml-good-at/index.html

img src="../fig/ep-03-cnn-diagram.png" alt="Wikipedia CNN Diagram"
Licence
CC BY-SA 4.0 Wikipedia
Proposed alt text: Diagram depicting layers of a convolutional neural network

img src="../fig/ep-03-titanic_newspaper_cv.png" alt="Titanic Newspaper CV"
Proposed Alt Text: An image of a newspaper front page comparing image classication with object detection.
Proposed caption: Image classification identifies that there is 90% probability the image is of a newspaper, and 70% it is the New York Tribune. Object detection identifies parts of the newspaper, such as nameplate and headline.

img src="../fig/ep-03-bird-collection.jpg" alt="USNM Bird Collection"
Proposed alt text: An overhead view of hundreds of drawers containing specimen trays of preserved birds of varying plumage, with one species per drawer.

img src="../fig/ep-03-pixplot-github.png" alt="PixPlot screenshot from GitHub"
Proposed Alt Text- A collection of images in PixPlot clumped by similarity,

Understanding and managing bias https://carpentries-incubator.github.io/machine-learning-librarians-archivists/05-managing-data-bias/index.html

Set of four wedding photos
img src="../fig/ep-04-Google_Inclusive_Activity_1.png" alt="Four separate images depicting wedding ceremonies"
7b) The same image as 7 but with descriptive labelling below.

Proposed Alt text: Four photos showing respectively 1) a man and a women in white clothes 2) a woman in a white dress, 3) a man in a tuxedo with a woman in a white dress, 4) a man in agbada and fila and a woman in wrapper, blouse and headpiece

Proposed Alt text: the same four images as previous image with the labels assigned by machine learning, demonstrating its biases.

**Text that follows image should read as follows:
The images are all from wedding photographs (donated by Googlers) for the Open Images dataset.
The classifier’s label predictions are:

  1. ceremony, wedding, bride, man, groom, woman, dress
  2. bride, ceremony, wedding, dress, woman
  3. ceremony, bride, wedding, man, groom, woman dress
  4. person people
    What do you notice about the classifier text for (4)?
    How well did you do compared to the classifier’s label predictions?**

Applying Machine Learning https://carpentries-incubator.github.io/machine-learning-librarians-archivists/06-applying-machine-learning/index.html

Notepad scribble of processing pages to machine readable text
img src="../fig/05-business-use-ocr.png" alt="Illustration of a business problem"
Proposed updates alt text- Diagram of workflow for using a paid for service to convert images to machine readable text.

Notepad scribble of text to catalogue process
img src="../fig/05-predictions-to-actions.png" alt="Two possible ways of using Machine Learning Predictions"
Proposed Alt text: Diagram of Workflow using machine learning to classify documents by language, with human oversight.

Notepad scribble of Digitized newspaper processes.
img src="../fig/05-model-choices.png" alt="Illustration of possible model choices"
Recommend splitting image and matching to bullet point order
Proposed Image A) Diagram of machine learning models which are ‘deeply’ embedded into applications
Proposed Image B) Diagram of Workflow using external closed source blackbox service.
Proposed Image C) Diagram of Workflow of building your own system using existing resource components.

Diagram of Workflow using an application that incorporates AI.
Notepad scribble of photograph to preservation process
img src="../fig/05-using-model.png" alt="A simplified GLAM workflow"
Proposed Alt text: Diagram of a simplfied GLAM workflow for publishing and preserving images.

@marionwalton
Copy link
Contributor

marionwalton commented Jan 31, 2024

I followed up on the Yoruba wedding alt text and found the correct names for the woman's attire as well

Four photos showing respectively 1) a man and a women in white clothes 2) a woman in a white dress, 3) a man in a tuxedo with a woman in a white dress, 4) a man in agbada and fila and a woman in iro ati buba with a gele.
https://carpentries-incubator.github.io/machine-learning-librarians-archivists/05-managing-data-bias/index.html

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
good first issue Good for newcomers
Projects
None yet
Development

No branches or pull requests

5 participants