In the Content Everywhere industry, AI and ML adoption varies greatly by company and its product or business model. The use of technology to process content is on the rise, says Better Software Group (BSG) CEO Bart Lozia. Deliver content to your audience. Solutions are more sophisticated, which means we can leverage them to our benefit and that of our users. “
Magnifi Chief Revenue Officer Meghna Krishna said: “Generative AI has become a popular topic these days with the advent of ChatGPT, but many brands have not considered the full potential of generative AI in areas such as video production. AI-based solutions that enhance workflows from brainstorming to editing.”
new trading tools
So who is doing what with AI and ML in the Content Everywhere industry? And how will the use of this technology evolve in the future?
At Ai-Media, as the name suggests, AI plays a key role in ensuring the accuracy of captioning, transcription, translation, and audio description services, for example. Ai-Media’s chief product officer, Bill McLaughlin, said advances in automated transcription and translation technology have “changed the way we work more radically than most other specialties in media production.” says.
“Compared to a decade ago, when professionally delivered content was rarely subtitled automatically, the volume of accessibility subtitling has increased exponentially, often at a price per hour of production. is down 90%,” he says. “Even if manual editing by experts is still involved, starting with AI-generated text as a guide will result in faster production times and lower costs than doing it entirely by hand. We’ve also made a lot of progress beyond just language models, such as AI/ML technologies for positioning and styling subtitles in , and generating generative textual descriptions of scenes and sound effects. “
At the same time, he cited quality control response as the biggest challenge facing the company as it adopts AI and no-touch workflows with customers.
“Human content editors and translators are not perfect either, but with automated systems there is a higher risk of being significantly off the mark. It’s easy to spend on monitoring and reviewing AI, which is why Ai-Media uses tools such as statistical quality measures to guide our approach and provide our customers with both cost- and quality-optimized results. I needed more experience in the field,” McLaughlin says.
Meanwhile, LTN SVP and Global Product Head Rick Young says his company is experimenting with AI “particularly for our high-volume, scalable production versioning service, LTN Arc. To drive the variety of production solutions we offer for a wide range of sports.”
In many cases, “there are significant efficiencies in allowing one operator to actively create two, four, or more versions of a live game,” says Young.
Looking ahead, LTN anticipates that many events will be able to run almost hands-free with operators assigned to respond only on an exception basis. This includes on-screen identification and translation of graphics, live-action events, in-game audio, or dynamic identification of scene changes and other data sources.
“While there has been much discussion about the risks of AI associated with the automated creation of misinformation and its potential for misuse, we believe that AI and automation will facilitate content creation and customization to meet needs. We’re focused on how we can bring targeted content to a wider audience – offering all kinds of relevant live programming for specific geographies, demographics and platforms,” added Young.
Some companies, such as Red Bee Media, have been using AI for years. According to Chief Technology Officer Richard Kidd, “Red Bee Media Access Services already had 20 years ago when he used AI speech recognition to generate live captions.”
Today, Red Bee Media says it uses Automatic Speech Recognition (ASR) throughout its subtitling workflow for both fully automated subtitling and improved workflow efficiency.
“We believe that accurate automatic speaker recognition is the newest frontier for fully automated captioning. Fully automated captioning for live captioning is growing rapidly as accurate delivery speed is very important compared to offline captioning.” has matured into something,” says Kidd.
He further adds: “The latest developments in large-scale language models are expected to further improve accuracy as these models represent a deeper context. , quality assessment, editing to increase reading speed, script extraction, etc. In addition, opportunities for further automating voice descriptions appear to be advancing rapidly, and sign language avatar technology appears to be on the horizon in the near future. Further progress is possible, but using avatars is still only suitable for certain scenarios.”
Planetcast International’s COO of Digital, Venugopal Iyengar, said the company’s use of AI/ML “has gone from lab experiments to deployment so amazingly quickly that AI is a top priority when tackling new projects.” It is now available,” he said. A recent example is Indian Premier League Cricket (IPL), his second-most expensive sports right in the world, which now uses AI to automate the creation of instant his highlights. “
Iyengar says AI is game-changing for Planetcast’s Contido suite of content supply chain management solutions
“We have been using AI-powered technical checks and baseline work such as segment/break recognition to improve throughput. We have also used AI in areas of compliance (such as language recognition). AI-driven features the Contido team is currently working on include using the Whisper speech recognition engine to auto-generate captions and aid in searching. We use AI to build custom NER (Named Entity Recognition) models to improve the accuracy of tagging assets and assets,” he explains.
As another example, Magnifi’s Krishna said the company’s AI video editors have been “trained on extensive data and have the ability to identify key moments and footage from sports matches using meta tags. “There are.” By specifying tags such as ‘Goals’, ‘Yellow Cards’, ‘Cristiano Ronaldo’, the editor can automatically generate clips of all the goals and plays of a particular match in less than 45 seconds. “
Other Magnifi features include an auto-flip feature, which allows brands to customize videos for different social media channel sizes, said Krishna.
Yang Cai, CEO and president of VisualOn, said the company has developed an Optima It said it launched a bandwidth-saving solution called Izer. .
“Applying AI and ML to video streaming, especially live video streaming, is still in its infancy,” adds Cai. “Object classification, recognition, and facial recognition are tasks that AI and ML will have to manage in the future, helping solve problems ranging from protecting privacy to improving the quality of user experience.”
Where next?
So what does the future of AI/ML look like in Content Everywhere, and how will it affect workflows, jobs, etc.? And companies are not biased towards ethics or AI algorithms. How do we address concerns about
“If you think AI might help, it turns out that it does in most cases,” comments Iyengar. “Better yet, we will soon know if AI can meet this challenge. We are excited about the challenges and opportunities, our experience so far shows that we shouldn’t expect human alone or 100% AI solutions to be perfect, human supervised AI is usually the most reliable We know we can provide a solution.”
Krishna says a small team can now produce video at the same speed as a traditional broadcaster, on many channels simultaneously, with the same or better quality and creativity.
“But this doesn’t mean a mass job loss in the industry,” she says. “On the contrary, AI tools can act as valuable assistants to video production teams, freeing up time to focus on the creative side rather than on technical or mundane tasks that can be easily automated. It also creates opportunities for human resources, requiring essential skills in an AI-driven environment.”
Frank Miller, Chief Technology Officer at Varnish Software, observes that ML and AI have great potential in the content creation and delivery ecosystem, but the right application is for each of these tools. It warns of the need to have a clear understanding of the capabilities and challenges of
“Importantly, all of these opportunities come with the caveat that you need to understand the algorithms and techniques to reliably produce worthwhile results. Deep learning on language models, which is analogous to the human brain with multi-layered neural networks that need to be trained and curated.Small perturbations during training can cause biases leading to strange behavior. There is,” he says.
As with other technologies, Miller adds: “Choosing the right tools, recognizing their capabilities and limitations, and applying them appropriately is the key to creating value. Values such as anomaly detection in are key to protecting the customer experience, but as long as we have a well-defined policy framework around each ML/AI model, we put it in the workplace.”
Synamedia Chief Technology Officer Tzvi Gerstl said AI continues to transform media production workflows with new features and greater automation of tasks across content curation, video editing, transcription and subtitling, and metadata tagging. , said to save time and resources.
“Although AI is known for its ability to quickly analyze large video data sets, the role of AI in content creation is growing, whether it is to improve video quality, add visual effects, or automatically generate content using synthetic media. I think it will have the biggest impact on the industry,” he says.
Gerstle concedes that while AI raises concerns for some, “we believe we are at a tipping point because of the new opportunities that AI brings to the television and video space.”
Red Bee Media’s Kydd says he has no concerns about the technical aspects and “thoughtful application” of ML and AI.
“But I’m wary of any technology where excessive expectations and unrealistic hype can become fashionable rather than functional,” he says. “We are currently seeing a rush towards AI, recalling the bubble of the past when it became difficult to separate what was truly groundbreaking and valuable from what was novel and interesting with no practical application. I will let you.”
“In my experience, instead of labeling something as AI/ML/LLM and hoping for success, the focus is on the value of the result and how that value becomes important to the customer,” Kydd said. should guess.”
