Categories
Computing

These are the 10 most-viewed YouTube videos of all time

Being popular is about the only thing the most-viewed YouTube videos have in common with their top-performing predecessors. Even though YouTube videos like Charlie Bit My Finger or Chocolate Rain went viral during the first few years of YouTube’s content, they probably wouldn’t be among the kinds of videos that go viral now.

In fact, children’s programming and music videos are now among the most-viewed content on YouTube. Music videos, in particular, have enjoyed great success on the streaming site and, until recently, had been the majority of the most-viewed videos in YouTube’s history. Music videos still account for a substantial portion of the top 10 most-viewed videos, however. If these view counts are anything to go by, the video-sharing site could be considered a leading platform for music videos and kid-friendly content rather than just the meme-worthy viral videos the site was known for in its early days.

What is the most-viewed YouTube video of all time?

Baby Shark Dance is the most-viewed video ever on YouTube. The children’s song overtook the all-Spanish version of Despacito in November 2020.

What are the top 10 most-viewed YouTube videos?

  1. Pinkfong — Baby Shark Dance (11.04 billion views)
  2. Luis Fonsi — Despacito featuring Daddy Yankee (7.93 billion views)
  3. LooLoo Kids — Johny Johny Yes Papa (6.40 billion views)
  4. Ed Sheeran — Shape of You (5.77 billion views)
  5. Wiz Khalifa — See You Again featuring Charlie Puth (5.59 billion views)
  6. Cocomelon Nursery Rhymes — Bath Song (5.54 billion views)
  7. ChuChu TV — Phonics Song with Two Words (4.76 billion views)
  8. Mark Ronson — Uptown Funk featuring Bruno Mars (4.64 billion views)
  9. Miroshka TV — Learning Colors Multi-Colored Eggs on the Farm (4.60 billion views)
  10. Masha and the Bear — Recipe for Disaster (4.50 billion views)

10. Masha and the Bear — Recipe for Disaster (4.50 billion views)

Маша и Медведь (Masha and The Bear) – Маша плюс каша (17 Серия)

Recipe for Disaster is one of many episodes from the Russian CGI-animated TV show for kids Masha and the Bear. It’s the only video on this list that isn’t a song (or features one song heavily) and only one of three that isn’t entirely in English. The show is widely popular outside its home country, and many episodes have huge view counts on YouTube. It’s also on Netflix in a collective episode format.

The Recipe for Disaster episode is of particular interest due to its astronomical view count.

Get Movies, the channel that uploaded Recipe for Disaster, earned loads of viewers just from this one specific episode.

9. Miroshka TV — Learning Colors Multi-Colored Eggs on the Farm (4.60 billion views)

Educational videos being among the most-viewed content on YouTube isn’t surprising if you think of it like this: If kids are going to be staring at screens for a decent portion of the day, it makes sense that the programming of choice among parents would be free to access and, better yet, educational. Miroshka TV’s Learning Colors Multi-Colored Eggs on the Farm is no exception. It’s an educational video for kids that has already racked up over 4 billion views in the four years since it was first uploaded.

It’s a fun, five-minute video that teaches young children about different colors via (crudely) animated eggs on a farm. The video itself is narrated in Russian and teaches kids the names of different colors in Russian. The only parts of the video that aren’t in Russian are the music (which is Old MacDonald Had a Farm in English) and any subtitles that appear (if you have them turned on). The video has a slow pace — some scenes are a bit long — but the funny, silly moments will appeal to younger audiences.

8. Mark Ronson — Uptown Funk featuring Bruno Mars (4.64 billion views)

Even though it’s older than the top three videos on this list, Uptown Funk continues to gather new viewers to remain a potential chart-topping threat. Ronson’s video made its YouTube debut in November 2014 and swiftly became one of the most popular and well-liked videos ever.

Arguably just as impressive as its view count is that it beat such stiff competition from its contemporaries. Big music videos from other major artists, like Katy Perry, saw their big 2014 hits achieve billions of views. Uptown Funk still manages to eclipse those achievements.

7. ChuChu TV — Phonics Song with Two Words (4.76 billion views)

It’s a pretty straightforward educational children’s literacy video. Phonics Song with Two Words features a simple tune that goes through the entire alphabet, going over pairs of words that use each of the letters in the English alphabet. The video even provides illustrations of these words and explains how to pronounce the sounds of each letter in their respective example words.

This video has been around since March 6, 2014, and has racked up over 4 billion views.

6. Cocomelon Nursery Rhymes — Bath Song (5.54 billion views)

Slowly but surely, educational videos for children are taking over the top 10 most-viewed YouTube videos of all time. In fact, Cocomelon’s Bath Song is now No. 6 on our list, as other videos like Gangnam Style have been knocked out of the top 10 completely– which is weird when you think about it. It’s just a simple song about bath time that’s less than three minutes long and has a Baby Shark Dance-esque chorus. And yet, it managed to snag its spot on this list, speeding far ahead of a video from an internationally famous singer.

5. Wiz Khalifa — See You Again featuring Charlie Puth (5.59 billion views)

Wiz Khalifa – See You Again ft. Charlie Puth [Official Video] Furious 7 Soundtrack

Wiz Khalifa and Charlie Puth’s See You Again is the first video on this list to have a movie tie-in — Furious 7 — to aid its promotion. It has managed over 5 billion views since its April 2015 debut and serves as a tribute to Paul Walker, the former star of The Fast Saga, who died in 2013.

Between July 10 and August 4, 2017, the video remained the most-viewed video on YouTube, dethroning Gangnam Style. That reign was only temporary, as it was later eclipsed by Despacito. It was also the most-liked video between August 26, 2016, and July 25, 2017.

See You Again became the second video to pass the 2 billion view mark.

4. Ed Sheeran — Shape of You (5.77 billion views)

Whether you watched it or not, there’s no denying that Shape of You is one of the most-watched videos of all time. It’s No. 4 on this list, has over 5 billion views, and was the second video to nab 4 billion views.

Shape of You debuted in January 2017.

3. LooLoo Kids — Johny Johny Yes Papa (6.40 billion views)

Another video for kids that made it into the top 10 most-viewed YouTube videos is Johny Johny Yes Papa. It’s a short, animated music video about a mischievous baby and his Papa. It’s a simple song about a baby trying to eat some sugar (literal sugar cubes) even though his Papa obviously doesn’t approve. Does the kid get the sugar? You’ll have to watch to find out. The song itself isn’t annoying, either — it’s almost too simple and short to be a nuisance. That’s a good thing, considering it’s been viewed/listened to more than 6 billion times.

2. Luis Fonsi — Despacito featuring Daddy Yankee (7.93 billion views)

Featuring Puerto Rican artists Luis Fonsi and Daddy Yankee, Despacito hit YouTube in January 2017 and racked up more than a billion views in just 97 days — the second-fastest run ever. The video went on to become the fastest to garner 2 billion views and the first-ever video to exceed 4 billion, 5 billion, 6 billion, and now 7 billion views.

1. Pinkfong — Baby Shark Dance (11.04 billion views)

Sometimes a song becomes famous solely because it’s an earworm. We’ve seen it happen with Barney in the early days, so it’s no surprise it happened again with an insanely catchy kid’s song about a shark family.

Baby Shark Dance, produced by Korean pop manufacturer Pinkfong, isn’t a complicated song. It has sharks, repetition, and a considerable amount of existential dread — that’s about it. However, this song rose to become the number one video on this list and has inspired countless variations found across YouTube.

Let’s be honest: Baby Shark Dance‘s success isn’t really all that surprising. If bringing up Baby Shark and playing it 12 times in a row is the only way to calm your toddler down, every parent with access to the internet is going to do exactly that. Expect Baby Shark‘s reign as the most-viewed video to continue.

Do you want to watch these videos without an internet connection? Read our guide on how to download YouTube videos.

Editors’ Choice




Repost: Original Source and Author Link

Categories
AI

Synthesia, which is developing AI to generate synthetic videos, secures $50M

Hear from CIOs, CTOs, and other C-level and senior execs on data and AI strategies at the Future of Work Summit this January 12, 2022. Learn more


Synthesia, a company leveraging AI to generate videos of avatars, today announced that it raised $50 million in a series B round, bringing its total raised to $66.5 million. Kleiner Perkins led the round with participation from GV, FirstMark Capital, LDV Capital, Seedcamp, MCC Ventures, and individual investors, which CEO Victor Riparbelli says will be put toward supporting growth and advancing Synthesia’s technology.

As pandemic restrictions make conventional filming tricky and risky, the benefits of AI-generated video have been magnified. According to Dogtown Media, under normal circumstances, an education campaign might require as many as 20 different scripts to address a business’ worldwide workforce, with each video costing tens of thousands of dollars.

Synthesia says its technology can reduce the expense to as low as $30.

“Synthesia is focused on reducing the friction of video creation and making it possible for anyone to create professional-looking videos in minutes, directly from their browser,” Riparbelli told VentureBeat via email. “Synthesia’s first commercial product, Synthesia Studio, launched in public beta in the summer of 2020. It is now used by thousands of companies, including several Fortune 500 companies.”

Generating synthetic videos

Like rivals Soul Machines, Brud, Wave, Samsung-backed STAR Labs, and others, Synthesia employs a combination of AI techniques to create visual chatbots, product demonstrations, and sales videos for clients without actors, film crews, studios, or cameras. Founded in 2017 by Riparbelli, Steffen Tjerrild, and computer vision researchers Matthias Niessner and Lourdes Agapito, Synthesia claims to have generated more than six million videos for over 4,000 clients — including SAP and Accenture — in the last year alone.

Synthesia

Above: A synthetic avatar created with Synthesia’s tools.

Image Credit: Synthesia

Synthesia customers choose from a gallery of in-house, AI-generated presenters or create their own by recording about 5 to 40 minutes’ worth of voice clips. After typing or pasting in a video script, Synthesia generates a video “in minutes” with custom backgrounds and an avatar that mimics a person’s facial movements and how they pronounce different phonemes, the units of speech distinguishing one word from another.

Synthesia says that client CraftWW used its platform to ideate an advertising campaign for JustEat in the Australian market featuring an AI-manipulated Snoop Dogg. The company also worked with director Ridley Scott’s production studio to create a film for the nonprofit Malaria Must Die, which translated David Beckham’s voice into over nine languages. And it partnered with Reuters to develop a prototype for automated video sport reports.

Synthesia recently made generally available a product that personalizes videos to specific customer segments. Aptly called Personalize, it can translate videos featuring actors or staff members into over 40 languages. Wired reports that more than 35 partners at EY, formerly Ernst & Young, have used Personalize to create what they call “artificial reality identity,” or ARIs — client presentations and emails with synthetic video clips starring virtual body doubles of themselves.

Synthesia

Above: Synthesia’s avatar creation dashboard.

Image Credit: Synthesia

“Our core use case today learning and development and internal communications videos, where the front-facing AI avatars work really well,” Riparbelli said. “The new investment will partly go to expand our core AI platform, which will allow more use cases.”

Deepfake concerns

Some experts have expressed concern that tools like Synthesia’s could be used to create deepfakes, or AI-generated videos that take a person in an existing video and replace them with someone else’s likeness. The fear is that these fakes might be used to do things like sway opinion during an election or implicate a person in a crime.

Recently, a group of fraudsters made off with $35 million after using forged email messages and deepfake audio to convince an employee of a United Arab Emirates company that a director requested the money. And just last month, Japanese police arrested a man for using deepfake technology to effectively unblur censored pornographic videos.

“[Deepfake technology is] becoming cheaper and more accessible every day … Audiographic evidence must be viewed with greater skepticism and must meet higher standards,” researchers in a new study on deepfakes commissioned by the European Parliament’s Technology Assessment Committee wrote. “[Individuals and institutions] will need to develop new skills and methods to construct a trustworthy picture of reality as they will inevitably be confronted with deceptive information.”

Synthesia

For its part, over-60-employee Synthesia says it has posted ethics rules online and vets its customers and their scripts. It also requires formal consent from a person before it will synthesize their appearance and refuses to touch political content.

“Our main competition is text — boring PDFs that people don’t read. Synthesia is driving a paradigm shift in how we create video content,” Riparbelli continued. “Looking into the next decade of Synthesia, we’re building for a future where you can create Hollywood-grade video on a laptop. On our way there, we’ll be solving some of the hardest and most fundamental problems in AI and computer vision. With the new funds, we’ll invest even deeper in advancing our core AI research to accelerate this vision. In parallel, we will also slowly open up some of our research to the world and begin actively contributing to the broader research community.”

VentureBeat

VentureBeat’s mission is to be a digital town square for technical decision-makers to gain knowledge about transformative technology and transact.

Our site delivers essential information on data technologies and strategies to guide you as you lead your organizations. We invite you to become a member of our community, to access:

  • up-to-date information on the subjects of interest to you
  • our newsletters
  • gated thought-leader content and discounted access to our prized events, such as Transform 2021: Learn More
  • networking features, and more

Become a member

Repost: Original Source and Author Link

Categories
Computing

How to Download YouTube Videos

YouTube has quickly become the world’s second largest search engine behind Google, with people spending hours per week on the popular platform. If you’re a YouTube fan like us, you might want to know what the most viewed YouTube videos are, and how to download YouTube videos. Also, check out our guide on how to download YouTube videos on an iPhone or iPad, if that is the device you will be using.

But first, we do need to issue this warning: Though it is legal to share and embed YouTube videos on the internet, downloading them for offline use isn’t always allowed. Downloading videos is considered a violation of  YouTube’s terms of service, unless the video streaming site has explicitly granted permission to download a particular video.

That said, let’s take a look at how to download YouTube videos. This guide features an easy-to-use desktop app known as 4K Video Downloader.

Step 1: Download the 4K Video Downloader app

screenshot

Head over to the 4K Video Downloader website, and under the heading 4K Video Downloader, select the blue Download button that corresponds to your computer’s operating system. This app is available for Windows, MacOS (10.13 or later), and Ubuntu. Once the installer has downloaded, run it to get the app installed on your computer.

If you don’t have a 4K display, don’t be put off by the name. This is one of the most versatile and simple-to-use download tools available, and it will work with all of your files regardless of your computer’s display resolution. The free version of this software has the ability to download individual videos at customizable qualities all the way up to 4K, has support for 3D and 360-degree videos, and can download subtitles.

There is a paid version that starts at $15, but for downloading a simple YouTube video, the free version will suffice.

Step 2: Copy the YouTube video’s URL

4k video downloader youtube url copy screenshot
screenshot

The next step is the easiest: Just head over to the YouTube video you want to download and copy its URL from the address bar at the top of your browser window. With the URL locked and loaded in your clipboard, you can close the window or tab of the video.

Further reading

Step 3: Paste in the URL

4k video downloader paste link screenshot
screenshot

Open up the 4K Video Downloader application you just installed. There’s no need to manually paste in the URL — just click on the Paste Link button in the top-left corner of the menu bar and the software will grab the URL from your computer’s clipboard.

Step 4: Download your video

4k video downloader step 4 download video screenshot
screenshot

The 4K Video Downloader will take a few moments to process the video. Once this process is complete, the app will let you choose from different video qualities, formats, and conversions. You’ll also be able to change where the video downloads; the application will construct a special folder for you containing all of your downloaded videos. If you use VPN, you should note that this specific software uses videos based on your IP address rather than the URL you copied. You’ll need to change your download location to the correct version.

Once you pick all your preferred options, click the Download button. Once you do this, a progress page will appear, detailing the size of your download, how fast it downloaded, and how much time remains in your download. You can pause the download if you need to without losing any progress. Once your download is complete, you can select your video from the open page. Look for the three vertical dots on the right-hand side and click them. A menu will pop out with the option to select Show In Folder. When you click this, the app will automatically take you to the file location of the downloaded video. 

After you’ve mastered it, you can configure 4K Video Downloader to automatically download videos or use Subscriptions or Smart Mode as necessary.

YouTube TV has taken off as well, you would be amazed to know how many subscribers YouTube TV has already. You might want to see if it’s worth subscribing to those shows.

Editors’ Choice




Repost: Original Source and Author Link

Categories
AI

SaaS-based film studio provides avatar ‘actors’ to make instant videos

Join gaming leaders online at GamesBeat Summit Next this upcoming November 9-10. Learn more about what comes next. 


If you’re dreaming of heading to Hollywood to make it big as a star, you can now add yet another reason to stay home and cling to that day job. A South Korean company named DeepBrain AI has announced the creation of AI Studios, a SaaS-based studio that makes it possible to create films without, as their media advisory spells out, the need to “film in-person or employ real people. ”

The unions for actors, directors, and film creators and directors might be in for a shock because of this.

To make matters worse for Hollywood workers and real estate holders, DeepBrain AI’s  product is available as a cloud service, so anyone can make a film and bypass California-level taxes by logging in from low-rent states that snarky Hollywood insiders used to call “flyover country.”

The tool features 30 new avatar-like “model/actors” that can, unlike almost all pretty faces from the casting agencies, speak fluent English, Korean, Chinese, and Japanese. The company promises that more languages will follow in 2022.

The system uses artificial intelligence to coordinate the motion of the actors with the script. The company promises that an aspiring director will only need to upload the text of the dialog — the AI will take it from there and synthesize the gestures of the arms or legs from a collection of preconfigured movements. There is currently no indication of whether the AI also handles the deep motivation and resonant backstory that fills out each character.

“With AI Studios, I was able to make a video using AI announcer Kim Hyun-Wook, a famous Korean announcer, without hiring, filming, or spending valuable time on the editing process,” said Cho Byung-hyeon, CEO of the Korean firm Commentary House.

SaaS-based film studio targeted for small and midsize businesses

While the development is a threat for actors at all levels of fame, DeepBrain AI is targeting individuals and smaller businesses. They expect that the AI models will find work adding a bit of life, albeit simulated, to what ordinarily might be dry PowerPoint or Prezi slides for YouTube videos or corporate training sessions.

DeepBrain AI is also partnering with real schools to put the AI models in front of students. The company put beta versions in Sinbong Elementary School in Seoul, Korea, where it was deployed in second and sixth grade. The teachers can share lesson plans and edit the presentations before they’re rendered for the children.

“Our AI tutors augment my lessons through video synthesis that mimics my teaching delivery as well as lesson reviews from other instructors, giving students additional information related to the lesson plan,”  said Hyungjun Yoon, one of the teachers at a Sinbong Elementary School.

Several samples of the AI tutors and are available on YouTube, including videos explaining how AI studios can ease video creation, how to utilize AI kiosks to amplify the next generation of customer engagement, and how AI human technology can assist with company clients and customer service. The current pricing plan will include three levels and a 20% discount for paying annually.

VentureBeat

VentureBeat’s mission is to be a digital town square for technical decision-makers to gain knowledge about transformative technology and transact.

Our site delivers essential information on data technologies and strategies to guide you as you lead your organizations. We invite you to become a member of our community, to access:

  • up-to-date information on the subjects of interest to you
  • our newsletters
  • gated thought-leader content and discounted access to our prized events, such as Transform 2021: Learn More
  • networking features, and more

Become a member

Repost: Original Source and Author Link

Categories
Computing

Microsoft Edge Version 92 to Stop Autoplay Videos By Default

Ever go to a webpage in Microsoft Edge only to have a video start playing in the background without your permission? Microsoft has heard those pains, and starting in Edge version 92, will change a settings toggle by default so that by default, you’ll no longer be annoyed.

The news is highlighted in the Microsoft 365 roadmap, as pointed out by TechRadar. In that road map, Microsoft mentions that Edge version 92 will help you keep your focus online by changing the default for auto-playing media to “Limit” from “Allow.” The page also mentions that the feature is rolling out now, so if you’re in the Beta or Dev and Canary beta channels and beta testing preview versions of the browser, you might already be seeing this change applied.

Although this setting will change by default in Edge version 92, you can still change the toggle on other versions of the browser (including the current version 91) manually right now. Just search for Media on Edge’s settings page, and then clicking the Media autoplay button at the bottom of the page. There will then be a drop-down box for Limit.

Google Chrome has had this setting option since 2018.  As Microsoft Edge is based on the same code as Google Chrome, it’s no surprise to see this setting option land in version 92.

This is just one of many features that have rolled out to Microsoft Edge since its initial release in early 2020. Microsoft has also added other features such as a new Kids Mode, vertical tabs, Collections, and even Windows 11 design elements. Even the performance has improved, with Microsoft introducing features like Sleeping Tabs, and Startup boost, designed to help the browser run faster and more efficiently — but also save memory.

And it’s all working in Microsoft’s favor. The old version of Microsoft Edge struggled with performance, was linked to Windows 10 featured updates, and had fallen behind with features when compared to Google Chrome. The new version of Microsoft Edge, however, has pushed forward, surpassing Firefox as the second most popular web browser in the world. It’s now on the same release cycle as Chrome and always seeing big improvements.

Editors’ Choice




Repost: Original Source and Author Link

Categories
Game

Pokemon GO Fest 2021 start times and rundown in kickoff videos

This weekend starts Pokemon GO Fest 2021, with all sorts of celebratory events and happenings therein. The game Pokemon GO started approximately 5 years ago, making this the official fifth anniversary event, and the biggest Pokemon GO event of the year. With the start of the event series, Niantic CEO John Hanke presented a video welcoming everyone to the party and another “podcast” was released by the company to start the event off right.

Below you’ll see the first “Pokemon GO Fest 2021: Event Kickoff” video with Niantic CEO John Hanke. This video gives users a quick overview of what the event will consist of and how the world of Pokemon GO is going right this minute. It’s been a long time since we called it – and it remains true today: Pokemon GO is the best game EVER: Here’s why (June 30, 2016).

Niantic also released a video with Niantic 3D designer Crag Kintzmann. This video shows a step-by-step process of constructing this year’s official Print at Home kit for Pokemon GO Fest 2021. This design was created in the first place by Niantic designer Ting Tey. You can find this kit at Niantic ready to download and print.

Finally there’s a Pokemon GO “podcast” of sorts from Niantic in the form of a video posted to YouTube. This video has a still frame of two Pikachu near the Pokemon GO Fest 2021 logo, onstage and ready for musical action.

Pokemon GO Fest 2021 has a start time of 10AM local time on July 17, 2021. This event ends at 6PM local time, also on July 17. The second half of the event will begin at 10AM local time on July 18, 2021, and will end at 6PM local time this same day. Take a peek at the timeline below for more information on the goings-on with this event and tips on how to succeed in this multi-day journey.

Repost: Original Source and Author Link

Categories
Tech News

YouTube on iOS PiP makes it much easier to watch videos while multitasking

YouTube is now rolling out the ability for all users to watch videos with picture-in-picture mode, which reduces the video players to a small floating screen on one’s phone or tablet. This feature won’t be limited to only premium customers as some had previously speculated, though those premium customers will get access to the PiP support first.

Android users have had access to YouTube’s picture-in-picture mode for a while; it has become increasingly useful as devices get larger, higher-resolution displays, leaving ample room for using more than one app at a time. With PiP, someone can watch a video in a small corner of their device’s display while doing something else, such as browsing social media, messaging, or playing games.

There have been concerns over recent months that YouTube wouldn’t only make its picture-in-picture mode available to paying Premium subscribers on iOS, but that’s not the case, according to confirmation given to MacRumors.

The feature is now rolling out to all iOS users in the United States, with Premium customers getting it first followed by free users ‘soon.’ Some iOS users have already had access to the YouTube picture-in-picture feature, though its availability has been touch and go with it working only sometimes.

The official support will eliminate the need to deal with difficult workarounds and buggy Safari streaming, though you may need to remain patient if you’re not a Premium user. The feature will be most useful on larger iPhone models where there’s enough screen space to watch a video in the mini player and engage in a second activity.

Repost: Original Source and Author Link

Categories
AI

This AI system learned to understand videos by watching YouTube

Elevate your enterprise data technology and strategy at Transform 2021.


Humans understand events in the world contextually, performing what’s called multimodal reasoning across time to make inferences about the past, present, and future. Given text and an image that seem innocuous when considered apart — e.g., “Look how many people love you” and a picture of a barren desert — people recognize that these elements take on potentially hurtful connotations when they’re paired or juxtaposed, for example.

Even the best AI systems struggle in this area. But there’s been progress, most recently from a team at the Allen Institute for Artificial Intelligence and the University of Washington’s Paul G. Allen School of Computer Science & Engineering. In a preprint paper published this month, the researchers detail Multimodal Neural Script Knowledge Models (Merlot), a system that learns to match images in videos with words and even follow events globally over time by watching millions of YouTube videos with transcribed speech. It does all this in an unsupervised manner, meaning that the videos haven’t been labeled or categorized — forcing the system to learn from the videos’ inherent structures.

Learning from videos

Our capacity for commonsense reasoning is shaped by how we experience causes and effects. Teaching machines this type of “script knowledge” is a significant challenge, in part because of the amount of data it requires. For example, even a single photo of people dining at a restaurant can imply a wealth of information, like the fact that the people had to meet up, agree where to go, and enter the restaurant before sitting down.

Merlot attempts to internalize these concepts by watching YouTube videos. Lots of YouTube videos. Drawing on a dataset of 6 million videos, the researchers trained the model to match individual frames with a contextualized representation of the video transcripts, divided into segments. The dataset contained instructional videos, lifestyle vlogs of everyday events, and YouTube’s auto-suggested videos for popular topics like “science” and “home improvement,” each selected explicitly to encourage the model to learn about a broad range of objects, actions, and scenes.

Merlot AI

The goal was to teach Merlot to contextualize the frame-level representations over time and over spoken words, so that it could reorder scrambled video frames and make sense of “noisy” transcripts — including those with erroneously lowercase text, missing punctuation, and filler words like “umm,” “hmm,” and “yeah.” The researchers largely accomplished this. They that in a series of qualitative and quantitative tests, Merlot had a strong “out-of-the-box” understanding of everyday events and situations, enabling it to take a scrambled sequence of events from a video and order the frames to match the captions in a coherent narrative, like people riding a carousel.

Future work

Merlot is only the latest work on video understanding in the AI research community. In 2019, researchers at Georgia Institute of Technology and the University of Alberta created a system that could automatically generate commentary for “let’s play” videos of video games. More recently, researchers at Microsoft published a preprint paper describing a system that could determine whether statements about video clips were true, by learning from visual and textual clues. And Facebook has trained a computer vision system that can automatically learn audio, textual, and visual representations from publicly available Facebook videos.

Merlot AI

Above: Merlot can understand the sequence of events in videos, as demonstrated here.

The Allen Institute and University of Washington researchers note that, like previous work, Merlot has limitations, some owing to the data selected to train the model. For example, Merlot could exhibit undesirable biases because it was only trained on English data and largely local news segments, which can spend a lot of time covering crime stories in a sensationalized way. It’s “very likely” that training models like Merlot on mostly news content could cause them to learn racist patterns as well as sexist patterns, the researchers concede, given that the most popular YouTubers in most countries are men. Studies have demonstrated a correlation between watching local news and having more explicit, racialized beliefs about crime.

For these reasons, the team advises against deploying Merlot into a production environment. But they say that Merlot is still a promising step for future work in multimodal understanding. “We hope that Merlot can inspire future work for learning vision+language representations in a more human-like fashion compared to learning from literal captions and their corresponding images,” the coauthors wrote. “The model achieves strong performance on tasks requiring event-level reasoning over videos and static images.”

VentureBeat

VentureBeat’s mission is to be a digital town square for technical decision-makers to gain knowledge about transformative technology and transact.

Our site delivers essential information on data technologies and strategies to guide you as you lead your organizations. We invite you to become a member of our community, to access:

  • up-to-date information on the subjects of interest to you
  • our newsletters
  • gated thought-leader content and discounted access to our prized events, such as Transform 2021: Learn More
  • networking features, and more

Become a member

Repost: Original Source and Author Link

Categories
Tech News

Camtasia 2021 is a down-and-dirty video editor that helps craft pro quality videos without the time and cost

TLDR: With Camtasia 2021, no-nonsense video producers can knock out a professional grade video that looks fantastic and won’t eat up hours of time to make, all at $100 off its regular price.

Video production. The words instantly make any content creator get anxious. They aren’t usually overly worried about creating the video itself. They’re worried about the time and money that go into such an event. 

While Hollywood blockbusters and the high cost of video production suites like the Adobe Creative Cloud have many fearful they’ll end up footing the bill for a $200 million James Cameron-style spectacular, there are still plenty of smart, even cost-effective ways of delivering a brilliant digital video quickly and painlessly.

Camtasia 2021 is ready to step in on that project, offering a host of options for making a professional-grade video for any presentation, social media post, advertisement, and more that won’t cost an arm and a leg.

Camtasia works with a very simple streamlined approach. Loaded with pre-built video templates, this software lets users record their computer screens, import PowerPoint presentations and basically add all kinds of visual aspects to their project, then turn it all into a video.

Without resorting to the fancy editing tricks of those ultra-expensive editing suites, Camtasia is a beginner-friendly platform and interface focused on easy solutions that make creating a video both quick to produce and attractive for viewers.

Using Camtasia, creators can record from their computer screen or from a webcam, then add high quality visual effects that fit with the project, all with simple drag-and-drop functionality. With a click, you can build in eye-catching titles and annotations, zoom in, pan across, animate objects and even transition between scenes with the skill of an Oscar-winning editor.

Castasia also includes a complete library of royalty-free music and sound effects for projects, so you’ll never run afoul of copyright infringement lawyers.

And as the 2021 edition, this version includes a whole collection of brand-new features, including 75 new and modern transition effects, motion blur and corner rounding abilities, and customizable media clips to go with your newly-shot video, logos, color schemes, and more. 

Right now, this bundle includes a copy of Camtasia 2021 as well as a year of program maintenance, which not only offers priority support and exclusive training, but will also hook you up with a brand new copy of Camtasia 2022 when it’s eventually released.

Regularly $299, this extremely limited time deal cuts that price by a third, getting you Camtasia 2021 and a year of app support for just $199. This offer is only available for a few more days, so get in on the deal now while you can.

Prices are subject to change.

Repost: Original Source and Author Link

Categories
AI

Facebook’s next big AI project is training its machines on users’ public videos

Teaching AI systems to understand what’s happening in videos as completely as a human can is one of the hardest challenges — and biggest potential breakthroughs — in the world of machine learning. Today, Facebook announced a new initiative that it hopes will give it an edge in this consequential work: training its AI on Facebook users’ public videos.

Access to training data is one of the biggest competitive advantages in AI, and by collecting this resource from millions and millions of their users, tech giants like Facebook, Google, and Amazon have been able to forge ahead in various areas. And while Facebook has already trained machine vision models on billions of images collected from Instagram, it hasn’t previously announced projects of similar ambition for video understanding.

“By learning from global streams of publicly available videos spanning nearly every country and hundreds of languages, our AI systems will not just improve accuracy but also adapt to our fast moving world and recognize the nuances and visual cues across different cultures and regions,” said the company in a blog. The project, titled Learning from Videos, is also part of Facebook’s “broader efforts toward building machines that learn like humans do.”

The resulting machine learning models will be used to create new content recommendation systems and moderation tools, says Facebook, but could do so much more in the future. AI that can understand the content of videos could give Facebook unprecedented insight into users’ lives, allowing them to analyze their hobbies and interests, preferences in brands and clothes, and countless other personal details. Of course, Facebook already has access to such information through its current ad-targeting operation, but being able to parse video through AI would add an incredibly rich (and invasive) source of data to its stores.

Facebook is vague about its future plans for AI models trained on users’ videos. The company told The Verge such models could be put to a number of uses, from captioning videos to creating advanced search functions, but did not answer a question on whether or not they would be used to collect information for ad-targeting. Similarly, when asked if users had to consent to having their videos used to train Facebook’s AI or if they could opt out, the company responded only by noting that its Data Policy says users’ uploaded content can be used for “product research and development.” Facebook also did not respond to questions asking exactly how much video will be collected for training its AI systems or how access to this data by the company’s researchers will be overseen.

In its blog post announcing the project, though, the social network did point to one future, speculative use: using AI to retrieve “digital memories” captured by smart glasses.

Facebook plans to release a pair of consumer smart glasses sometime this year. Details about the device are vague, but it’s likely these or future glasses will include integrated cameras to capture the wearer’s point of view. If AI systems can be trained to understand the content of video, then it will allow users to search for past recordings, just as many photo apps allow people to search for specific locations, objects, or people. (This is information, incidentally, that has often been indexed by AI systems trained on user data.)

Facebook has released images showing prototype pairs of its augmented-reality smart glasses.
Image: Facebook

As recording video with smart glasses “becomes the norm,” says Facebook, “people should be able to recall specific moments from their vast bank of digital memories just as easy as they capture them.” It gives the example of a user conducting a search with the phrase “Show me every time we sang happy birthday to Grandma,” before being served relevant clips. As the company notes, such a search would require that AI systems establish connections between types of data, teaching them “to match the phrase ‘happy birthday’ to cakes, candles, people singing various birthday songs, and more.” Just like humans do, AI would need to understand rich concepts comprised of different types of sensory input.

Looking to the future, the combination of smart glasses and machine learning would enable what’s referred to as “worldscraping” — capturing granular data about the world by turning wearers of smart glasses into roving CCTV cameras. As the practice was described in a report last year from The Guardian: “Every time someone browsed a supermarket, their smart glasses would be recording real-time pricing data, stock levels and browsing habits; every time they opened up a newspaper, their glasses would know which stories they read, which adverts they looked at and which celebrity beach pictures their gaze lingered on.”

This is an extreme outcome and not an avenue of research Facebook says it’s currently exploring. But it does illustrate the potential significance of pairing advanced AI video analysis with smart glasses — which the social network is apparently keen to do.

By comparison, the only use of its new AI video analysis tools that Facebook is currently disclosing is relatively mundane. Along with the announcement of Learning from Videos today, Facebook says it’s deployed a new content recommendation system based on its video work in its TikTok-clone Reels. “Popular videos often consist of the same music set to the same dance moves, but created and acted by different people,” says Facebook. By analyzing the content of videos, Facebook’s AI can suggest similar clips to users.

Such content recommendation algorithms are not without potential problems, though. A recent report from MIT Technology Review highlighted how the social network’s emphasis on growth and user engagement has stopped its AI team from fully addressing how algorithms can spread misinformation and encourage political polarization. As the Technology Review article says: “The [machine learning] models that maximize engagement also favor controversy, misinformation, and extremism.” This creates a conflict between the duties of Facebook’s AI ethics researchers and the company’s credo of maximizing growth.

Facebook isn’t the only big tech company pursuing advanced AI video analysis, nor is it the only one to leverage users’ data to do so. Google, for example, maintains a publicly accessible research dataset containing 8 million curated and partially labeled YouTube videos in order to “help accelerate research on large scale video understanding.” The search giant’s ad operations could similarly benefit from AI that understands the content of videos, even if the end result is simply serving more relevant ads in YouTube.

Facebook, though, thinks it has one particular advantage over its competitors. Not only does it have ample training data, but it’s pushing more and more resources into an AI method known as self-supervised learning.

Usually, when AI models are trained on data, those inputs have be to labeled by humans: tagging objects in pictures or transcribing audio recordings, for example. If you’ve ever solved a CAPTCHA identifying fire hydrants or pedestrian crossing then you’ve likely labeled data that’s helped to train AI. But self-supervised learning does away with the labels, speeding up the training process, and, some researchers believe, resulting in deeper and more meaningful analysis as the AI systems teach themselves to join the dots. Facebook is so optimistic about self-supervised learning it’s called it “the dark matter of intelligence.”

The company says its future work on AI video analysis will focus on semi- and self-supervised learning methods, and that such techniques “have already improved our computer vision and speech recognition systems.” With such an abundance of video content available from Facebook’s 2.8 billion users, skipping the labeling part of AI training certainly makes sense. And if the social network can teach its machine learning models to understand video seamlessly, who knows what they might learn?

Repost: Original Source and Author Link