Skip to content

Poet image description tool: Now with enhanced training module, EPUB 3 support

2014 October 14
Submitted by Sue-Ann Ma, Product Manager, Benetech
Benetech released an updated version of Poet, our open-source image description tool, on Friday, September 26th. The highlight of the update is our new Image Description Training Module. The new training module introduces novice describers to the importance of providing accessible images and leverages important research from DIAGRAM Center partners to provide more user-centric tools that help describers figure out “When to describe?” This first phase would not have been possible without the subcontract research work from Touch Graphics on their Decision Tree and online templates that guide novice image describers. Concepts from the Accessible Image Sample Book from the DIAGRAM Content Working Group were also integrated into the module. Feedback from users has been very positive. As one veteran “image slammer” from a Benetech university partner program put it:
“The new training is easily found all on one section of the POET website. It is short, concise, aesthetically pleasing and easy to understand. I love it!!!” - Rachel Schwartz, SIP Internship Director, Brigham Young University
The latest release also expands file format support so that image descriptions can now be created on Poet for EPUB3 files (previously limited to DAISY). Descriptions are tagged in accordance with industry-standard specifications, also noted as item #7 on the DIAGRAM Center’s “Top Tips for Creating Accessible EPUB3 Files.” Additionally, Poet’s updated Image Checker tool (login required) makes it possible for users to do a quick review of various types of image descriptions (such as alt, captions, aria-describedBy, aria-describedAt, longdesc) embedded in an EPUB3 file.
Benetech looks forward to growing the new training module, incorporating additional reference guides, best practice examples, and hands-on exercises with the continued involvement of the DIAGRAM Center community. Please feel free to take it for a spin and send us feedback. We welcome any and all comments and suggestions.

What We Are Reading

2014 October 30
by jnoblitt

Here are a few items that caught our attention in the last month:

    • Wonderful TEDx talk by Chelsea Cook, a blind physics student from Virginia Tech who talks about how 3D printing made complex math concepts clear to her for the first time, including some great in-classroom footage of the teacher-student interaction where she has her “aha” moment – very moving: “Touch the stars”

Tactile Graphics with a Voice wins “Best Student Paper” award

2014 October 30

Earlier this year we reported that Richard Ladner’s DIAGRAM research was completed on a tool that converts image labels into a QR code that can be affixed to a graphic, allowing the image to be read and voiced by a smart phone. A paper has now been published about the work, which is part of the Tactile Graphics Project at the University of Washington. The paper is available as open access until early October 2015 and after that it is closed in the ACM Digital Library. Retrieve it here:

Catherine M. Baker, Lauren R. Milne, Jeffrey Scofield, Cynthia L. Bennett, and Richard E. Ladner. 2014. “Tactile graphics with a voice: using QR codes to access text in tactile graphics.” In Proceedings of the 16th International ACM SIGACCESS conference on Computers & accessibility (ASSETS ’14). ACM, New York, NY, USA, 75-82.

The paper was awarded Best Student Paper at the ASSETS 2014 conference this month. According to Dr. Ladner, a version of the iPhone app TGV to help read QR-codes will be in the App Store soon, and an updated version with a finger pointing feature should be in the store in a couple of months.  Congratulations to Richard and his team!

Evolution of a 3D Campus Map

2014 October 30
by jnoblitt

Submitted by Lore Schindler, Los Angeles Unified School District

Last year, we bought our first 3D printer to test out how we might produce tactile representations to connect our students with visual impairments to the world of visual images. I quickly realized that the potential of 3D printing for our students is huge, but that the software to create the images involved a learning curve. We needed help.

Fortunately, last summer we met Joan Horvath, an engineer working with Deezmaker, (the 3D printer store where we bought our printer), and the author of Mastering 3D Printing. She connected us with a wonderful group of young students at Pasadena City College (PCC) and their Fab Lab. I asked if they would create a tactile map of our elementary school campus, and they started to work immediately. These college students were amazing! In three weeks they walked our school campus, learned the basics of the braille code, designed and printed a lightweight and foldable tactile map of our school campus. I showed the map to some of our Orientation and Mobility instructors. They were impressed with the design and suggested a few changes to fit how they orient their students to the campus. Within a week, the PCC students responded with a version 2.0 of the map—we were blown away by their responsiveness and dedication to the project. I asked one of our students who reads braille to explore the map, and his response after some hands-on time was, “It looks cool and I wish I had one.” Check out his initial investigation of the 3D campus map from this YouTube link.

While the PCC students have moved on to other projects, including creating 3D images of chemical molecules, we’re continuing to work with Deezmaker and hope to have a 3D map of one of our middle school campuses completed soon.

September 2014 Advisory Board Meeting

2014 October 30

The DIAGRAM Center Advisory Board held a quarterly meeting on September 22, 2014. In addition to the reports from each of the Working Group co-chairs, the Board heard updates on Benetech’s new IMLS award as well as a new accessibility working group that will be chaired by Robin Seaman for the Book Industry Study Group (BISG). Sue-Ann Ma and Bryan Gould joined the discussion as well to talk about the new Poet Training Module which incorporates refreshed NCAM image description guidelines and image description templates developed by Touch Graphics. Notes have been posted in the community area of the DIAGRAM blog. NOTE: You must be registered and logged in as a DIAGRAM Community member to see the notes. If you have any questions, please write to Julie Noblitt, Community Manager, at julien [at] benetech [dot] org.

September 2014 Content Working Group Update

2014 October 30

The DIAGRAM Content Working Group met on September 22, 2014. Topics included continued discussion of planning for the Accessible Image Registry/Repository, the revised and expanded image description guidelines from NCAM, and the new accessibility working group for the Book Industry Study Group (BISG) to be chaired by Robin Seaman. Notes from the meeting have been posted. The next meeting’s discussion is expected to focus on a proposed revised edition of the Accessible Image Sample Book. If you have any questions about this Working Group please contact co-chairs Elaine Ober <Elaine [dot] ober [at] pearson [dot] com> and Lucia Hasty [lucia [at] tactilegraphics [dot] org> or visit the wiki page to see past meeting notes.

October 2014 Software Development Update

2014 October 30
by jnoblitt

Poet Upgrade Released

Read all about the latest enhancements, including a brand-new training module and support for EPUB 3 file formats!

Tools Working Group Update

The DIAGRAM Tools Working Group met on October 8, 2014 for a special presentation by Fred Esch (Accessibility, Watson Innovations AARB Complex Visualization Working Group Chair) and Tom Babinski (IBM Human Ability and Accessibility Center) who demonstrated RAVE, a charting component that produces accessible charts.  The presentation was recorded. You can hear and see the demo by visiting this link (access is free; you will be prompted to register but your information will not be shared – it’s for DIAGRAM’s internal tracking only). Refer to the Tools WG wiki for past meeting notes and contact the Tools Working Group Chair Geoff Freed at <Geoff [underscore] Freed [at] wgbh [dot] org> if you would like more information.

October 2014 Training & Outreach Update

2014 October 30
by jnoblitt

Next DIAGRAM Webinar: Haptics!

Title:                Integrating Haptic Feedback to Enhance Learning Materials
Date:                TBD
Presenter:      Markku Hakkinen, Educational Testing Service

What is haptics, you ask? It’s a super-cool tactile feedback technology that simulates the sense of touching physical features on a flat, tablet screen.  Don’t miss this session with haptics researchers Markku Hakkinen of Educational Testing Service (ETS) and others as they update us on new ways that haptic feedback can enhance learning materials. Hear the results of new DIAGRAM-funded research that builds upon existing research at ETS in tablet-based haptic display of graphical information and extends it to explore the inclusion of this technology to eTextBooks.

Editor’s note: this post originally indicated a webinar date of December 11, but it will be rescheduled for early 2015.

Outreach Working Group Update

The DIAGRAM Outreach Working Group met on October 9, 2014 to discuss the refreshed strategic goals for 2015. At center stage was a new goal for the group, determining metrics of success so that we can more meaningfully articulate our impact. We are delighted that for our next quarterly meeting in December (date TBD) DIAGRAM program evaluator Maureen Hawes will join the discussion to give us her insights on this important topic. Contact Outreach Working Group Chair Julie Noblitt at julien [at] benetech [dot] org if you would like more information or have ideas for discussion. See the DIAGRAM Outreach Working Group wiki for past meeting notes.

DIAGRAM Out and About: October 2014 Update

2014 October 30
tags: , , , ,
by jnoblitt

October 2014 Standards Update

2014 October 30
by jnoblitt

@longdesc moves toward W3C recommendation

Submitted by Geoff Freed, National Center for Accessible Media, WGBH

The inclusion of the longdesc attribute in HTML5 has been debated for nearly five years, but at long last the matter has been settled. This week the W3C resolved a formal objection to the HTML5 Image Description Extension (also known as the longdesc extension to HTML5), clearing the way for the document to continue moving toward a full W3C recommendation. You can read the full ruling at . It’s a very long decision but is very much worth reading. The W3C’s decision and, eventually, approval of the extension as a formal recommendation will give authors full support for using longdesc as one method among several to provide detailed image descriptions.

For those who like to read the end of mysteries first, here’s the final ruling:

On the basis of careful consideration of all points raised, the Director does not find sufficient basis to uphold this formal objection and it is hereby overruled.

The Joint HTML Accessibility Task Force, and the two Working Groups under which the Task Force operates, may therefore continue to progress this specification per W3C Process.

We encourage all parties to apply their creative energies and design capabilities constructively towards improved solutions for accessible image descriptions in the future.

W3C on HTML5

Submitted by Geoff Freed, National Center for Accessible Media, WGBH

The W3C has formally advanced HTML5 for full recommendation status, as they announced in a statement this week: . While it isn’t news that HTML5 and related technologies such as SVG and MathML (which are included within the full specification) have been in use for some time, the recognition of the markup as a formal W3C recommendation means that it’s a complete standard, at least until HTML5.1 is issued sometime in 2015. For DIAGRAM, it might mean that SVG and MathML may start replacing their image counterparts in textbooks and other materials.

Standards Working Group Update

The DIAGRAM Standards Working Group continues to meet jointly with the TIES (Transition to Inclusive EPUB 3 eco System) Working Group on a biweekly basis, every other Tuesday. The groups have made great progress on enhancements to the DIAGRAM Content Model, and have made progress with a braille rendition in the IDPF. Please contact George Kerscher at <kerscher [at] montana [dot] com> if you would like more information. All meeting notes continue to be posted on the DIAGRAM Standards WG wiki.

Benetech Receives IMLS Grant for National Forum on 3D Printing for Accessible Education

2014 October 9
by jnoblitt

We are extremely excited to announce that the Institute of Museum and Library Services has now awarded Benetech a 2014 Laura Bush 21st Century Librarian Program grant to hold a National Forum on the intersection of 3D printing, libraries and museums, and accessible education. This year-long project seeks to define how existing and upcoming 3D printing programs in libraries and museums across the country can help overcome technology and resource challenges so that 3D printing can be applied to improve STEM education for students with disabilities. The National Forum, tentatively scheduled for Spring/Summer 2015, will:

  • bring together 30+ experts in these fields to survey and understand existing efforts
  • design an effective partnership model with key stakeholder groups, and
  • propose effective metrics for additional projects in this area.

In addition to holding the forum, we’ll create a few resources, including a Quick Start Guide for educators and maker spaces, and a small collection of 3D printable models optimized for classroom use.

The next steps for this project will include putting together an Advisory Committee that will help us work closely with the library, museum, and maker communities as well as set the agenda for the National Meeting. We hope to engage many of you soon in the discussions ahead. We are excited to grow our community of stakeholders and keep pushing for better ways to make sure all students get the educational materials they need. As always, stay tuned!

DIAGRAM Out and About: August 2014 Update

2014 August 26
tags: , , ,
by jnoblitt

Tobi 2.4 Brings Parallel Production Capability for EPUB 3 Projects

2014 August 26
by jnoblitt

Submitted by Varju Luceno

In June 2014, the DAISY Consortium announced the new release of the next generation open source multimedia book production tool, Tobi. Tobi is a unique software tool that can produce publications compliant with both the DAISY 3 and the mainstream EPUB 3 specifications. It includes the unique capability of adding accessibility enhancements, such as image descriptions, to the mainstream EPUB 3 publications. From the Change Log:

DIAGRAM image descriptions: EPUB3 export (XHTML5), including audio and alternative images (like DAISY projects), based on @longdesc, @aria:describedAt, @aria:describedBy / iframe + fixed metadata issue when importing DIAGRAM XML.

This release also addresses reported issues and includes other bug fixes. To download Tobi or to learn more about it, please visit the Tobi project area For more information about what’s new in this release, please visit the change log

DAISY Pipeline 2 Version 1.8 Brings Enhanced File Conversion Options

2014 August 26
by jnoblitt

Submitted by Varju Luceno

The DAISY Pipeline 2 is a cross-platform, open source tool for converting documents into accessible formats for people with print disabilities. The latest release, Version 1.8 includes a variety of bug fixes and improvements, including the new Text-to-Speech (TTS) based DTBook to DAISY 3 converter, better handling of image descriptions, rewritten HTML5 converter as well as the production of EPUB 3.0.1-compliant publications.  For details, please see the release notes on Github. (Pipeline 2 releases are hosted on GitHub servers since version 1.7.)

The HTML to EPUB 3 converter intends to implement the DIAGRAM Center’s “Image Guidelines for EPUB 3” by converting any local DIAGRAM descriptions to off-page HTML descriptions embedded as visually-hidden iframe elements and linked to images by aria-describedby attributes.

The first preview of the newly developed Text-to-Speech based audio production modules and Natural Language Processing-based structure detection modules is included in this release, available in a new DTBook to DAISY 3 converter. Text-to-Speech based production will be further tested and improved over the coming months to support the production of EPUB 3 with TTS-based Media Overlays from DTBook or HTML. TTS adapters are available for a variety of platforms and TTS engines including Acapela TTS (v7), eSpeak, SAPI 5 voices on Windows, and Mac OS X voices. Developers are working on a simple Graphical User Interface (GUI) that will be available in the end of 2014.

Pipeline 2 Version 1.8 requires the installation of Java SE 7. Also, please note that the SAPI 5 adapter currently requires the installation of Visual C++ Redistributable Package’s runtime components, available for free on Microsoft’s web site.