Harvard and M.I.T. Are Sued Over Lack of Closed Captions
nytimes.com
nytimes.com
Similar discussion as previous thread on UC Berkley: https://news.ycombinator.com/item?id=12519761
If the universities avoid releasing free videos, that punishes all of us including the majority of people who don't have disabilities. I'd prefer to find another way to caption these videos besides throwing around lawsuits.
If a Federal appeals court rules that Netflix doesn't have to caption their commercial videos[1], surely we can cut some slack for free education videos.
[1]http://arstechnica.com/tech-policy/2015/04/9th-circuit-rules...
Disability advocacy groups have pursued this cause through other channels, to no effect. This lawsuit is a last resort. The institutions have had a long time to get it together; as the article says, a letter was sent by the DOJ way back in 2010.
No, not all. I was restricting my argument to things given away for free with no commercial gain.
As society, we've agreed that making commercial entities like restaurants install wheelchair ramps is good.
However, if we see a soup kitchen feeding the hungry poor and notice that the building they use doesn't have a wheelchair ramp, we should threaten them with a lawsuit?!? (Yes, the hungry poor can be wheelchair bound too.) It seems means-spirited and a misguided use of the ADA.
There are several lines of attack on this argument. The first is against your contention that these institutions enjoy no commercial gain. There are clear PR gains from offering these courses, both through engendering goodwill and increasing prestige. I am not saying these courses are offered only for PR reasons, but it is difficult to deny their effect.
The second line of attack is against your belief that free goods & services should not be subject to accessibility requirements. You use the example of a soup kitchen. Here I will digress for a moment and say I believe analogies are an exceptionally poor form of argument in civil rights contexts, which operate on abstract principles such as the value of equality. Thus an emotional appeal to our intuition about what is mean-spirited is unlikely to be useful. Returning to the argument at hand, we can question what is special about a free good or service which enables its exemption from accessibility requirements while remaining consistent with our principles of equality. So, I pass the ball back to you - what is your answer for this question?
Our law provides many special cases when the entity is engaging in non-profit activities: taxes, labor requirements, among others.
So you ask why non-commercial should entities be excluded from accessibility requirements: because they already get excluded from other onerous requirements we place on businesses, and it seems like a "reasonable" line to draw.
Maybe you don't find it reasonable, but now it's up to a court decide.
> the majority of people who don't have disabilities
Almost 19% of the US' population is disabled in one way or another. That's a lot of folks to discriminate against. But you're right, they're the minority, so what rights should they have to get access to publicly available services? Rip out those ramps, get rid of the parking spaces, close up the isles, and shut down those elevators.
> Netflix doesn't have to caption
Netflix won that appeal, and lost the next. Rather than take it further, they made a deal and added closed captioning. So, yes, if you offer public content, it should be available for everybody. Cost of doing business.
While this content may be nominally free, it's an advertisement for MIT and Harvard classes. As much money as these universities spend on their regular marketing materials, they can afford a few TAs to add captions to their video courses.
It'd be great if they were closed captioned, true. But it's already "good enough" that they're public. Giving them even a moment's doubt about the value of putting their knowledge in the public is a bad idea.
A disservice to society is treating some people who, through no fault of their own, as worth less than everyone else.
> Giving them even a moment's doubt about the value of putting their knowledge in the public is a bad idea.
So you'll allow them to hold their data hostage to get special favors and exceptions from federal law? Where do you draw the line then?
They already provide wheelchair ramps, braille (and low vision) compatible signs, ARIA tags on their websites, door openers, all for free for anybody who comes to their campus; what makes their videos exempt?
That they have gotten away this long is them not paying attention to the laws; these lawsuits are simply an attempt to get them to follow the laws in the country they live and work in.
If these laws go through, Harvard and MIT should just set up a small subsidiary in another country and upload the videos through those. Problem solved.
https://www.insidehighered.com/news/2016/09/20/berkeley-may-...
So rather than helping anyone, you're sabotaging many.
I would like to point out though that merely uploading a video to YouTube results in it being captioned, for free, by a machine learning algorithm.
If it was a teacher that put his content up on his personal website, nobody would care. But because it was done in MIT's name, people start throwing dirt.
I've asked some hearing-impaired friends what they thought about this article. Their reaction was that it was ridiculous to sue over the accessibility of information given away for free.
Well, perhaps its because that material was paid for in part using public grants - your tax money and mine? These aren't poor, barely solvent community colleges, these are private institutions with huge revenues and investment funds in the hundreds of millions.
> If it was a teacher that put his content up on his personal website, nobody would care
If it were created with money from a federal grant, then yeah, people would still care.
Do you have any sources that shed light on whether any of this material is the product of federal grant money? You assume it is, when the article and other sources I can find offer no suggestions or evidence that the universities are releasing federally funded information in an inaccessible format. Based on your comments here, you have made a set of assumptions and concluded that because the universities accept federal funding, and because they have released course material to the public, that this course material is the product of federal money. That has not to my knowledge been shown to be the case. If you're including loans/grants that are awarded to students, then all colleges, even the poor, little, barely solvent ones, are to be subject to the same principles. But when you make what sound like snide remarks about an institution's investment funds and revenues, you somewhat betray that you're only outraged because the university has money, and isn't doing what you think it should with that money.
I originally didn't word that clear enough to avoid misinterpretation. The "net negative" was specifically about the free education videos and not about all ADA lawsuits in general. I added the "[about the free learning videos]" to make that reference clearer.
>Netflix won that appeal, and lost the next.
I see no cites that says Netflix lost another appeal to NAD after April 2015.
MIT and Harvard were not following the law; the lawsuit simply attempts to bring them back in line. Free videos or not. MIT and Harvard both provide multiple ADA allowances as part of the cost of doing business; I think the question should be why are they trying to get around this one? The ADA has a history of trying to resolve these issues directly before resorting to lawsuits.
> I see no cites that says Netflix lost another appeal to NAD after April 2015.
My mistake; they were approached with a new lawsuit, and immediately decided to settle; the settlement document from April of this year:
http://www.adatitleiii.com/wp-content/uploads/sites/121/2016...
I don't know -- I understand that it must not be helpful for deaf students, but it seems like it would be cheaper for the schools to just stop sharing the courses instead of dealing with a lawsuit. What does the NAD expect to have happen?
Indeed. That's exactly what happened at UC Berkeley, once news of these lawsuits got out. We used to have all of our class lectures available free to the public, but now they're restricted only to students.
Furthermore, the fact that these considerations are not at the forefront of people's minds when releasing materials reinforces the merits of the lawsuit and the law.
It's not "clear" at all that there's value to put it up for them. They probably did it out of good-will, in which case the easiest option was to simply record the lecture. I don't think it's fair to imply malice at all.
So the solution is for them to give you a call, and you'll do it for a cut of the profit (i.e. free), right?
I didn't imply malice at all.
No, they should not call me. They should start with an inclusive perspective that eliminates the need for lawsuits. And when things slip through the cracks because of advancements in technology, or cultural shifts in perspective they should lead.
My point is that it's embarrassing because they had the opportunity to do the right thing, and instead they did the easy thing, which was to take it away.
> they had the opportunity to do the right thing, and instead they did the easy thing, which was to take it away.
Nope, by providing it for free they definitely were adding. As you will see when they take it offline, it was completely additional.
EDIT: I see you were talking about UC Berkley, I though "they" referred to Harvard and M.I.T, my mistake. Yeah, they took it away, but lawsuits have that effect. I doubt the plaintiffs would have settled for "we'll crowdsource the captions" :)
> They should start with an inclusive perspective that eliminates the need for lawsuits
IDK. It's taken millennia to get to where we are. Improvement isn't an option? A campaign to crowdsource the close captions might have been far more effective and definitely less mean-spirited than a lawsuit.
The all-or-nothing approach isn't practical and will likely result in nothing.
I'm not sure what you're saying here, "Nope, by providing it for free they definitely were adding. As you will see when they take it offline, it was completely additional."
I think your crowd-sourcing idea is a good one. I think that would be great, but in this instance it is circuitous.
The simple reality is that in the US, we have agreed as a society that lawsuits are how we settle things. That is how we improve things.
It's not a negligible cost to transcribe every single phrase the lecturer utters, for hours and hours and hours of lessons.
Would it be better if the captions were generated by Youtube? Come on, that system is not good enough yet.
My school has an annual operating budget in excess of $100MM. It's a small school in the Midwest. They could slice something off to accommodate the closed-captioning.
I worked in post-production in Los Angeles and Manhattan and I have a reference for closed-captioning service fees. They could do it, or even better they could hire a person (maybe more) full-time and save money.
A technology solution that improves access for everyone sounds like a great starting point for a startup.
It's worth mentioning that captioning services are provided for students with disabilities, but that infrastructure simply isn't a part of what was made public, and is only provided in those classes that have students with disabilities in them.
It's one thing for them to provide public access to exactly the same portal that students used to access the webcasts that are being made anyway, but an entirely different one to also a). Run the risk of expensive lawsuits on the university's dime, and also b). ensure that those videos are up to federal accessibility standards.
Edit: wording and fixing dictation errors.
That's like saying that if a store donates usable meat, they have to give a vegetarian option.
All it accomplishes is to punish those who go through the effort to make sure that byproducts of their mission can reach more people, and make it harder to "recycle" such products.
Think about it this way: it's already legal for me to make blind-incompatible blog posts and videos and give them away for free. This law is saying "hey prestigious institution, it's okay to give away your course materials for free, but you better durn well funnel it through a no-name blogger so no one knows where to find it!"
/s
Original discussion: https://news.ycombinator.com/item?id=9039798
You don't sue a guy doing good deeds just because he is not 100% good to all, or missed someone in his efforts. Gosh.
EDIT: For online course material that is offered to enrolled students and required for the completion of a degree, I do not believe anyone is arguing in opposition.
First, let's examine the virtue of accessibility. We start with the observation that nearly 1 in 5 Americans are disabled[0]. The source provides a breakdown by disability type. 1 in 5 is a high enough ratio to appeal to purely selfish reasons for the virtue of accessibility: there is quite a good chance you yourself will become disabled and require accommodation, so you should advocate for accessibility while you are still fortunate enough to be healthy. Other reasons to support accessibility include improved economic output by including the disabled in the workforce, ensuring a path to success for exceptional disabled people, increased quality of life among the disabled populace (about which you may or may not care), and an appeal to the abstract ideal of equality.
Now that we have established the virtue of accessibility, let's consider the virtue of enshrining accessibility requirements in law with penalty for non-compliance. Clearly, not all services desired by the disabled will implement accessibility (an example is the topic of this thread). Of those services which do not voluntarily implement accessibility, does the government have any interest in compelling implementation? Your answer to this question will depend on your own beliefs about the role of government; to head off an enormous digression, we'll just assume that yes, the government does have such an interest.
If the government has an interest in compelling implementation of accessibility, to what services does this interest extend? Certainly not all; it would be absurd to require every favor or interaction between individuals maintain full access compliance. For our purposes here let's use a classification function unlikely to be controversial (and more restrictive than is practiced): services provided by the government, and services provided by institutions receiving money from the government.
Both MIT ($460m in 2015[1]) and Harvard ($656m in 2012[2]) receive funding from the federal government. Thus they satisfy our restrictive classification, and so services offered should implement accessibility measures subject to penalty. These free online courses are services offered by MIT and Harvard. Thus they should implement accessibility measures. They certainly have the means.
[0] https://www.census.gov/newsroom/releases/archives/miscellane...
[1] http://web.mit.edu/ir/financial/re.html
[2] http://www.thecrimson.com/article/2012/12/30/sequestration-r...
- freely publishing course materials to the internet-at-large is the same thing as providing educational services for which the universities receive federal funding
- services provided by institutions receiving money from the government includes all services provided--pending your clarification on how releasing information equates to a service provided--entirely, without exemption, and not just those services to which federal funds are directly allocated
- applying federal compulsion to meet desired accessibility goals should apply to individuals who are not the official recipients of the benefits of the programs funded by the federal government (in this instance, meaning the general public of internet users who are non-students desiring to consume freely available information)
I find this automatic set of claims/assumptions to be highly dubious, as it equates releasing information--even if organized by originating course, thus giving the appearance of being an "online course", but actually lacking any involvement or service from the university beyond releasing the material, and thus not really an "online course"--to providing education to enrolled students (for which the university receives funding and is inarguably subject to accessibility restrictions with associated penalties). I think these are, if one does not ignore the nuances of the circumstances, quite different things that weaken the argument for considering them to be services provided by the university. How do you enshrine this in a principle that applies to all actors? Does everything a university sponsors or engages in fall under these principles? Meaning, because they received federal funding for programs A, B, and C, do programs X, Y, and Z fall under federal purview and penalty, as well? Should the internet public really be granted the latitude to expect they can have their cake and eat it, too, where consuming course material for personal edification as a non-student is concerned?
You can check my much lengthier thoughts elsewhere in the thread on this, if you'd like. Also, let's exclude your "They certainly have the means" comment. That, to me, diminishes the force of your argument for principles by concluding it as an argument that the alleged principles only matter because the affected party has the perceived means. I think that severely weakens the case for why this isn't just a grab to hold an entity to a higher standard because they have the money to pay for the burden.
You raise a possible distinction between free online courses, which you reduce to releasing information, and services. This might be delving into semantic argument territory. I would say the onus is on you here to argue why free online courses should not be considered a service offered by the university and thus possibly not subject to accessibility requirements.
Edit: Please forgive posting a link to another comment. I'm pointing to the other comment only because I feel it is quite thorough, and am not sure the conversation is helped by duplicating the same sentiments, thoughts, and questions.
Edit 2: I hope my consideration of the issues and perspectives can pass the smell test of invoking a semantic argument. I genuinely find this particular issue to be one that is nuanced and important, and do not wish to subvert reasoned consideration by quibbling too much over words. Suffice to say, I think this is far more a matter of situational and conceptual difference, not just whether we should rightly use a word to indicate a thing.
1) Disability rights are often portrayed as a special privilege for a subgroup of society. But in reality, all of us start our lives mobility impaired and illiterate, and many of us who are lucky enough to grow old become sight, hearing, and mobility impaired to some extent. Accessibility helps all of us.
2) Accessibility helps not just humans. High quality captions add a lot of value to videos, because they make the videos searchable, which provides immense benefits to all users, hearing impaired or not.
3) Disability rights are the one aspect of social policies where the USA might be ahead of much of the rest of the world. Compared to my native Switzerland, the USA is far ahead in empowering disabled individuals to participate in public life. This is something the US should be proud of.
Thinking about what's easier if the choice is take down all material and not spend money on a lawsuit or spend even more money to add captions there is an obvious choice for universities to make :-/
Do they get a pass because it's not their content? Do they get a pass because it's almost impossible to provide? Or do they suck it up and CC everything at great expense to everyone?
I mean, that would be equivalent to suing a gallery because paintings are "discriminatory". Or the painter. Obviously ridiculous.
Additionally, providing inclusive access should be a point of pride and a highlight of the community and culture that executes it.
> Under Title III, no individual may be discriminated against on the basis of disability with regards to the full and equal enjoyment of the goods, services, facilities, or accommodations of any place of public accommodation by any person who owns, leases, or operates a place of public accommodation. Public accommodations include most places of lodging (such as inns and hotels), recreation, transportation, education, and dining, along with stores, care providers, and places of public displays.
[...]
> The statutory definition of "readily achievable" calls for a balancing test between the cost of the proposed "fix" and the wherewithal of the business and/or owners of the business. Thus, what might be "readily achievable" for a sophisticated and financially capable corporation might not be readily achievable for a small or local business.
It's a 25 year old law. I don't have much sympathy for large organisations that are still choosing to discriminate against disabled people.
It is clearly, unambiguously, discrimination. It might be legal (for the reasons I quoted around practicability), but that legality does not stop it being discriminatory.
But then you do a 180 and say it's wrong because it's immoral.
Obviously, it could be both, but even if the first argument is kind of a "duh" it seems more cogent than going back and forth between those two positions (as many people seem to do in the thread).
Society decided it was immoral to discriminate against people with disabilities. We tried asking nicely, and got ignored, and so we passed laws to drive change. Those laws were made some time ago, and while it takes time for change to happen we've had long enough for people to bring things into compliance with those laws.
This is especially true for large well funded organisations producing new material. There's not much excuse for a large well funded organisation, who is well aware of reasonable adjustments, to not make new material accessible.
Arguments against making material accessible are "it's too hard" or "it's too expensive". That's covered by the law (I included reference to it in my post).
But the fact that we still don't have useful speech-to-text or good transcription shows the contempt much of the tech industry has for people with disabilities. See also the choice to exclude people by choosing pale grey text on a bright white background. (Note that this doesn't just exclude people with disability levels of eyesight loss, but also people with mild visual loss).
"Harvard and MIT both requested a Motion to Stay or Dismiss the cases on the grounds that current accessibility law doesn’t explicitly require universities to caption video on the web.
In June, the US Department of Justice weighed in on the issue in a statement of interest to the court. The USDOJ argued in favor of the NAD and urged the judge to deny the Motion to Stay of Dismiss. Judge Robertson agreed.
That means that, barring an objection from the District Judge, the lawsuit will proceed without delay.[1]"
[0] http://www.3playmedia.com/2015/06/26/usdoj-sides-with-nad-in... [1] http://www.3playmedia.com/2016/02/23/judge-denies-motion-to-...
Sample scenarios:
- Color red is banned because colorblind people can't see it
- "screeching nails on blackboard" Youtube video is pulled because deaf people can't enjoy it
- Public datasets are pulled because the numbers are too complicated for ordinary people to understand
IMO there should be a general divide between "public data" and "public accommodation". Data is just bits; 0's and 1's can be communicated to anyone. Requiring them to be something else as well is going too far. To put it another way: the accessibility problem is in the human-computer interface, not in the service on the other end of the network.
And here's Harvard's brief agreeing with me, categorizing the videos as "goods" rather than "services": http://www.plainsite.org/dockets/download.html?id=235493653&...
That being said, it seems like the NAD only expects this to apply to public accommodations, so perhaps all that's needed is to move (donate) the videos to a separate (non-public-accommodation) location, e.g. Bittorrent.
It's like if an organization is giving away food and then gets sued for not providing a gluten or nut-free option.
Its the sentiment that if some people can't have it, no one should.
Yeah, it's not like MIT and Harvard have spare money lying around to spend on captioning (apart from the $13.5B and $37.5B, respectively, in their endowments). Let Big Deaf use their deep pockets to pay for captioning, if that's what they want!
To restate my final point: there is nuance here, and people are missing it. I'm commenting for the purpose of interrogating the nuance, because I feel somewhat mixed on the issue. For online course material that is offered to enrolled students and required for the completion of a degree, I do not believe anyone is arguing in opposition. However, the collection and publishing of course material provided to enrolled students, then sharing it for free to non-enrolled students is a different matter entirely. The free part is an important detail in this particular matter and its circumstances, and shouldn't be wholly ignored, or treated as if it isn't part of the equation.
> If the lectures were a dollar would you say that they're obligated to subtitle them then?
Possibly. Most likely, only if paying for the material and following the courses was somehow tied to earning enrolled status and credit toward a degree, though. Because at that point, someone is actually a student looking to obtain something in exchange for studying the material, the university is engaged in the activity for which it receives federal funding, and we would rightly expect the institution to treat them as students according to the law and all its glorious regulations that seek to provide all students with a level playing field. Giving the material to the public at large with no fees or strings attached--meaning no strings attached to either party--isn't something I think we should discourage.
> Or do you have a specific cost threshold before you expect subtitles?
There is no cost threshold on my mind, no. There is only the threshold of whether the parties consuming the materials are enrolled students seeking a degree at the institution.
> ... but they're also depriving disabled people of them ...
I'm not convinced this is true. The internet is full of freely available information from a variety of sources, much of it in video and audio form, and we do not have a longstanding debate centering on how much of the freely available information in video and audio form is depriving the hearing impaired of that information and should be made accessible. What's happening here is singling out a particularly easy target and asserting that they should be held to a different standard than all the other parties producing free, inaccessible content, and calling it "depriving disabled people" of the content. This stirs my something-isn't-quite-right detector, because we are attempting to provide a very narrowly scoped requirement onto a narrowly scoped party, on the basis of taking rules that inarguably apply to their services in one particular set of conditions, and applying them to another, quite different set of conditions.
> Also, are people forgetting that these are Harvard and MIT? The NAD isn't going to be suing your mom and pop website. Harvard and MIT can afford it and should be held to higher standards.
This is an argument from a pretty low set of standards, honestly. The ability of the party to afford increased accessibility sets up a rather disingenuous cash-gate on the issue, and completely debases the argument for accessibility into an argument about money. We're either concerned about establishing a proper set of guidelines and cultural expectations for making information accessible, regardless of its cost, or we're targeting entities with cash who are otherwise doing something we applaud, and saying because they have the means to do more, they should do more, and bringing the force of the state against them to compel them to do so. This is the kind of thinking that inexorably leads to crafting laws that target specific parties, leave open loopholes for other parties, and wind up subverting our intended goals by allowing those who wish to avoid a particular set of regulations and obligations by reorganizing under an uncovered entity type. We'd surely want to avoid such an outcome--even if it would help us better identify truly bad actors.
> I hate to use the word due to the anti-SJW frenzy the internet is in these days, but the ableism in this thread is appalling. No one is trying to see from the side of the NAD, with one user even suggesting that it just wants to line its coffers...
I don't think there is an appalling level of ableism in this thread. I am, and I think others are, trying to interrogate the issue from multiple perspectives, but we are coming to different conclusions (or are withholding conclusions) than you seem to expect. Perhaps that's because we're looking at the nuances of the circumstances.
From the NAD's perspective, I'm questioning and considering the affordances one ought to expect from information being made freely available in its original form, and what limitations can sensibly be agreed to exist--because it is insensible to expect there to be no limitations. This includes interrogating the alleged principles involved, and to what extent and to which parties they apply. When they don't apply equally to all parties, especially when they don't apply equally on the basis of one's ability to pay, I find the alleged principles reveal themselves to be suspect. This perspective, in particular, is the one in which the absence of affordances and obligations on information-releasing parties are felt most acutely. My lacking of a particular ability preventing me from equally enjoying informative and enlightening material is a bitter pill--especially if I can reasonably expect otherwise.
From the university and academia perspective, I am questioning and considering what reasonable thresholds one ought to be able to easily identify when releasing information in is original form freely to the public, or withholding it because additional accessibility affordances cannot reasonably be provided in light of the return on the time invested versus simply releasing the information. This perspective, in particular, is the one in which the force and burden of the obligations we levy as a society are felt most acutely. For instance, if a professor is teaching a course in which no persons enrolled have a disability, is the professor obligated to only use information which is accessible to serve the unknown contingent of internet consumers should the university decide to release the course materials freely on the internet? Are the lines only drawn at choosing videos with accurate captions? What about all the millions of people with other learning disabilities of some sort--what is the university obligated to do to ensure they are not "depriving disabled people" of this information? If other kinds of learning disabilities do not merit such affordances, why not? Why only this one or that one?
From the social and cultural perspective, I am questioning and considering what expectations and obligations we ought to hold in such cases for both parties, and how we should reasonably define these expectations for accessibility to as many people as possible and the obligations of implementation. Do we base our expectations and obligational determination on defining thresholds of sheer number of people who may potentially be affected by the lack of affordances? Do we only care about certain accessibility affordances, while ignoring others? Why or why not? We have, I think, passed the point of solving many of society's issues with hammers and saws. We now need scalpels. Much as medical science drastically improves outcomes by isolating bad things and eradicating them with precision, instead of simply removing a whole appendage, we need to pay attention to the nuances and rationally interrogate them to figure out what we think is best socially and culturally. If that's increasing the reach of federal disability law to cover information that is given away for free for the masses of internet consumers, okay. But we better establish some bulletproof and sane principles for doing so, and hold all information producers equally accountable. If we don't, then let's drop the veneer that we are holding all individual and organizational entities equally responsible and accountable, and admit we are instead targeting specific entities based on their perceived ability to pay for the increased obligation to be universally accessible.
Aside from feeling mixed in sum of all the above perspectives and not jumping to immediate and simplistic conclusions that ignore the nuances of circumstance, I continue to feel mixed because I think, as a principle, an accessible web is a better web. I hold firm to the principle that the more information people have access to, the better off they are, and the better a society is for providing this information as accessibly as possible to as many people as possible. However, I think there is also a somewhat disappointing need to include in our rational calculations when information producers, whomever they may be, are publishing that information as they have it, to put it out there, to share it widely with as many people as they can in the form it exists. Perhaps if we had better tools, we could abstract away the burden, then try taking the route of expecting entities to use certain technologies that alleviate the need to take on making things accessible on their own. If we make accessibility a social and cultural good and goal, how might that change how we produce information?
It would be great if everyone could afford to offer accessible content.
Maybe, instead of paying instructors, all lectures should be typed verbarim - in advance - and delivered by Text-to-Speech software (with gestural scripting and intonation). All in the same voice.
- Ahead-of-time lecture scripts could be used to help improve automated speech recognition accuracy.
- Provide additional support for paid captioning
-- Tools
-- Labor
- Provide support for crowdsourced captioning services
-- Feature: Upvote to prioritize
-- Feature: Flag as garbled
- Develop video-platform-agnostic transcription software (and make it available for free)
-- Desktop offline Speech-to-Text
-- Mobile offline Speech-to-Text
-- Speaker-specific language model training
- Require use of a video-platform with support for automated transcription
-- YouTube
--
- Companies with research in this space:
-- Speech Recognition, [Automated] Transcription, Autocomplete hinting for [Crowd-sourced] captioning
-- IBM
--- YouTube has automated transcription
--- Google Voice supports transcription corrections, but AFAIU it's not speaker-specific
-- Baidu
-- Nuance (Dragon,)
--
... Textual lecture transcriptions are useful for everyone; because Ctrl-F to search.
- Label (with RDFa structured data) accessible content to make it easy to find
-- Schema.org accessibility structured data (for Places, Events)
--- https://github.com/schemaorg/schemaorg/issues/254
--- http://schema.org/accessibilityFeature
--- http://schema.org/accessibilityPhysicalFeature and/or
--- http://schema.org/amenityFeature
--- https://github.com/schemaorg/schemaorg/issues/254
- Challenges
-- Funding
-- CPU Time
-- Error Rate
-- Mitigating spam and vandalism
-- Human-verified crowdsourced corrections can/could be used to train recognizing and generative speaker-specific models
-- In the film A.I. (2001), there's a scene where they're asking questions of Robin Williams and the intonation/inflection inadvertantly wastes one of their 3 wishes / requests. https://en.wikipedia.org/wiki/A.I._Artificial_Intelligence
That's not what we are talking about.
[Solutions for solving the problem (providing transcripts of lectures) most cost-efficiently]
- Microsoft
-- "conversational speech recognition"
- Apple
You can get the material anyways from the recommended course books, which are probably available in the public library at request. Or the course notes in the course's web page.
I agree with your point, however speech recognition likely resulted in the mentioned "inaccurately or unintelligibly captioned" videos ;)
Since then I am a strong supporter of the ADA.