Continuing the review of video trends in 2021, if you missed part one you can read it here. Part two is here. In part three the following topics will be covered:
- 5G Technology
- Ultrafiber 2,5GB
- Live video streaming
- Low latency
- Mobile first
- Multiview video streaming
- Educational Technology
- Learning Coding
- Video summarization
- Artificial intelligence and machine learning in video
The companies/projects talked about will be:
- 5G consortium
- Multiview Media
- Codingal
- Clipr
11. 5G Technology
Speed of light video delivery to mobile devices
5G is the upcoming next big improvement in mobile internet connection, and beyond. With shorter range but 10x the speeds of 4th generation networks, 5th Generation networks will bring unparalleled speed in mobile internet capabilities. It is a massive operation because it includes numerous stakeholders which have to provide the hardware infrastructure (masts, cables, antennas, 5G enabled mobile devices, etc), the software infrastructure (network switches, redundancy, communication operativeness, network management, etc) and the business use cases, like low latency high quality video streaming, virtual and augmented reality or 360 videos, which require incredibly high amounts of data, throughput and speed to be experienced correctly. Also, in todays geopoliticized world, security of the infrastructure and information that goes through it is paramount to national security, so there is an added layer of scrutiny on the security practices, encryption methods around 5G implementation and the actual people who will own the hardware and software stacks and hence be able to eavesdrop on the data if a malevolent actor. 5G has been rolling out at different paces throughout Europe and the World. In the UK for instance it is already present in many cities provided by E3 , and an ever increasing area is getting coverage. In other EU countries, such as Italy, big telco companies like Vodafone are giving out grants to companies bringing sound use cases using 5G and other technologies, like virtual reality. As a point in case, the 5G consortium brings together many of the actors involved in making this technology a reality for todays internet users and creating new exciting uses for technologies which are already available but require higher bandwidth for wider adoption. One of such use cases is video and immersive and volumetric video, which will bring the users to a new era of interactive content.
Link to project: https://www.5gsolutionsproject.eu/about/consortium/
12. Multiview Media
Multi camera interactive video streaming
This British startup is a great example of another video trend which the COVID19 pandemic has greatly accelerated due to the requirements of social distancing and non crowded events. The great times of hugging your friend after a goal standing up in a crowded stadium or singing along with thousands other while stomping your feet in a concert seem long gone, and nobody knows for sure how long this new normality of staying individually will continue for. As a new de facto reality throughout the world, that of not gathering together with other people has created a vacuum for sporting and artistic events which used to be crowded with many many people and now feel much emptier. Albeit TV has introduced the concept of distant viewing, the absence of people in stadiums and arenas has opened opportunities to get creative about enriching the viewing experience. This is what multiview video, and Multiview Media in the specific do, hence allowing the user watching the program to side step production and choose whichever camera, from multiple available, to stream the event from. This is more engaging for the user as it is interactive and caters the need to watch the event from a different angle, as close as you could do in person by turning your head or changing location and getting a different view of the event. Technically this can be done in the browser with multiple video player elements, so that there is 1 player which is used as the reference time for the event and the other streams, whether in HLS or DASH, representing the different cameras are always encoded and available and the containing video player can be selected with a click. Once the user clicks on a particular camera, the manifest will be loaded in the player and then that camera will be shown. The switching can happen faster by preloading the video manifest url in the player and having it already parsed and available once the user clicks on a camera. Also, using the master time reference from the main player, a newly selected camera if its time is off can be seeked to match the master time, hence not making the user lose any of the event. So this company, Multiview Media is using the technology to facilitate content providers and producers in delivering a superior experience to their users and allowing them to enjoy the event at a greater level than before.
Link to the company: https://multiviewmedia.co.uk/
13. Codingal
Educational Technology video and coding
The Indian based YCombinator backed startup represents a boosting trend in todays COVID era, educational technology or edtech. India especially has seen a number of new startups dealing with this topic of experiencing learning digitally mainly through video. With much (unfortunately) of the schooling of kids and teenagers shifted online, the edtech market has seen a huge increase in the past year, and the usage of readily available tools like Google Meet, Zoom or Microsoft Teams, have helped continuing as good as possible the knowledge sharing between students and teachers. With the over usage of videoconferencing services a highway of opportunities has opened to iterate on the available services and seek improvements. One of the companies innovating in this space is Codingal which creates a platform to facilitate kids and teens learning to code with video conferencing, tests, quizzes, lessons with the unique selling point being personalized, one-to-one learning between student and teacher, so that the speed of learning and breadth of topics covered is tailored to the specific student which can then come together with other classmates through the community and coding challenges and tests. The hope is that edtech coding platforms will increase to deliver the needed knowledge to fill the millions of tech jobs open and coming up.
Link to the company: https://www.codingal.com/
14. Clipr
Artificial Intelligence aided video summarization
This startup is the best one to describe an increasing trend in video technology, video summarization through Artificial Intelligence. The company founded by ex AWS Head of Computer Vision and other media experts focuses on summarizing a video to extract salient points and highlights. Traditionally, video summarization has envisioned selecting the IDR frames, or Interdependent frames, or frames that can be decoded by themselves and don’t need any previous or next frames to get the “full picture”. As this approach indeed extracts main parts of the video through a succession of key frames, it does not apply any intelligence to it. This is where deep learning and machine learning come in. Clipr uses underneath the APIs exposed by AWS Rekognition and its developing its own models to use multiple parameters and not only keyframes but also speech to text, object recognition in video, face identification and scene detection. These features together are helping content creators automatically create highlights of long form content and get it distribution ready without the lengthy manual post production and editing process. Many tech giants have entities identification in their cloud services offer (think Azure Cognito, Google Video AI, AWS Rekognition and so on) yet a dedicated platform for video summarization clearly defines a sub market segment which is eager to have summarized, searchable, entities identified video content. In the sports domain, DAZN is already using video summarization to automatically generate highlights of football matches and have them readily available as the game is happening and right after. So a service like Clipr is tackling a somewhat different (and harder) task because it is not domain specific and its general purpose service, to actually summarize any video has to know a lot more contexts – and here the power through AWS services shows – to understand and efficiently and correctly summarize multiple kind of videos. General purpose video summarization services like Clipr.ai will be increasing in the future, along domain specific services to automatically generate summaries of videos through Artificial Intelligence.
Link to this company: https://www.clipr.ai/