Dissemination of 4K videos, how can I handle the heavy capacity?


There is one thing in common when we look at the characteristics of recently released electronic devices. It is that it is equipped with a ‘4K resolution display’. Thanks to this, the current 4K displays are rapidly becoming disseminated. In addition, YouTube, the world’s largest video platform, has high-definition standards of 4K, and the public is getting used to 4K video.

As the use of video content has soared, the video market is literally booming, and with the development of technology, the spread of 4K video can be seen as a natural phenomenon. However, this phenomenon may not be welcomed by content providers. This is due to the ‘capacity’ of the video content.

Is expensive equipment the best way to create excellent quality content?

As the demand for content increased, the supply has increased. In the midst of the abundance of content produced in a day, creators are fiercely competing to provide better contents. In addition to the dense composition of the story, the quality of the sound and screen has now become an important factor in the selection of content. That’s why creators around the world tend to produce video content in 4K. This is because 4K resolution videos are more competitive among numerous videos, which can increase demand.

As the video market becomes active, it is said that the demand for equipment and software needed to produce content is increasing. This is because most content creators are trying to provide better quality videos by building better production systems. But for 4K images to be produced, it costs more than what’s normally used. It can be made with expensive products such as 4K and 8K camera, HEVC real encoder, and equipment that enables 3D sound.

If it is difficult to produce 4K video content due to the burden of equipment, upscaling can also be used. The quality of the HD image improves to UHD quality when the content is filmed conveniently with the existing equipment, and the upscaling is done after editing. In addition, kokoon.cloud, which provides with an online upscaling service, enables 4K upscaling simply by uploading without having to install a separate program, allowing anyone to use it without any burden.

A creator who produced 4K video, but the capacity is a problem?

However, even if expensive equipment is purchased or upscaling is used to create 4K content, another concern arises for creators. The capacity of video content will take up much more capacity than existing image content, and creators whose videos are professionally produced may feel burdened by the production due to the rapidly growing capacity of the video.

Most of the units of storage used today are terabytes (TB). However, terabytes have been lacking recently, the biggest cause of this phenomenon is ‘video.’ In particular, the capacity of 4K video that takes less than 30 minutes is close to 20GB, which can be considered to be almost 1GB per minute, so 60GB has to be given when a one-hour video is produced. In response to this, although the burden is slightly reduced with the advent of SSD and HDD, professional creators with a large number of contents that need to be produced regularly will feel the limit of storage space in a short period of time.

However, the lack of capacity does not mean that 4K resolution can be abandoned. Already, 4K video content has become disseminated to the public, and the level of viewing has increased, so image quality is important in order to gain a competitive edge in content. Especially if you’re a professional creator, you can’t give up on any of the elements that make up the content, the sound and image quality of the content. Therefore, what is needed most now is a solution that can make the capacity lighter as the image quality remains the same.

A magical solution that reduces data capacity by half while maintaining image quality

BLUEDOT has introduced ‘PQ-Optimizer’, a technology that optimizes cognitive image quality that reduces data while maintaining image quality. The self-developed AI is designed in such a way that the image is analyzed and preprocessed, so it can be applied regardless of the codec, and it finds and removes pixels that are unnecessary for the implementation of the quality of the image, which makes the difference in image quality almost not visible, allowing the data to be light. Thanks to the increased efficiency of the encoder compression, it solves the difficulties of lack of data capacity of video providers.

Conversely, if the usage of data becomes the same, it can implement higher picture quality. In addition, because less data is used, high-quality images can be reliably transmitted even if live streaming is done on an unstable network. Therefore, ‘PQ-Optimizer’ is a solution that allows creators whose videos are produced to be useful in various ways.

With the surge in demand for video content, the quality of video content has improved at a fast pace and is still ongoing. And there are many ways appearing to improve the quality of the video. With improved equipment, high-quality video content can be produced, but the use of the solution introduced in today’s posting can also be seen as an efficient production method to improve the quality of the video.

*Please contact the email below if you want more information about the ‘PQ-optimizer’ introduced in this post.

Contact: info@blue-dot.io

From ‘Video Restoration’ to ‘Face Application,’ how far have you used upscaling?

A movie “A Moment To Remember” photo upscaling Before-After (NAVER movie)

As the number of people enjoying video content increases, it is ‘technology’ that follows on its own as the market grows. As competition for content intensifies, creators often develop strategies to appeal to the public with higher image quality as well as solid story to provide higher-quality content. Therefore, there will be a demand for technology that improves image quality, and new technology will emerge.

Typically, there is ‘upscaling’. Pixels are created to enhance the resolution of images so that they can be useful when you want to watch them on a variety of displays, or when you want to improve the quality of images. Moreover, once the upscaling technology is applied, it can be used more widely. Among the technologies of various applications, ‘Remastering’ and ‘Face enhancement’ will be introduced today.

‘Remastering,’ which adds modern visual beauty to those days and the mood.

Leslie Cheung in a movie “Days Of Being Wild’ (NAVER movie)

Recently in the media industry, there is a growing craze for ‘remastering.’ At the end of last year, the “Remastering Project” was announced at SM Entertainment, saying that the video of a past-affiliated singer’s music videos and performances would be upscaled 4K and released, and the tour of the last concert of Leslie Cheung, a global star from Hong Kong, was restored to 4K video and released. And the public is showing enthusiastic support for the media’s actions, where past images have been restored to good quality and released.

Remastering is a technology that adds color, removes noise, and focuses on sharpening to restore the image clearly. And when it is gone through the upscaling phase, it is reborn as a complete modern image. Although upscaling is not a major technology, thanks to the addition of pixels, quality images can be produced to satisfy consumers.


In addition, as Cyworld, the SNS of memories, comes back, the contents of the past, including images, are being re-examined. So that each person can be active through SNS, the contents of that time and then are shared and enjoyed with people. When ‘remastering,’ a modern technology, is applied, then its mood is maintained and can be upgraded to video content with modern visual beauty added to it.

‘Face Enhancement’ applied to the video of ‘Face Application’ which was applied only to the photo

‘Face Enhancement’ is literally a technique for improving images focused on facial expressions. In addition to the skin tone, the outline and shadow will be added to make the face more vivid, so it will play a role similar to the ‘Face Apps’ used in the photos. Previously, the function was applied only when Face Apps were turned on and filmed, but with the advent of Face Enhancement, pre-shooting and post-production are now possible. Thanks to this, it is possible to improve the blurriness and low definition that have been the disadvantages of camera apps in mobile.

In fact, upscaling alone can improve basic tone correction and clarity. However, if it is an image that the person wants to improve depending on the type of image, if Face Enhancement is applied, you can receive the result that is improved in the desired direction that standing out the person, and if upscaling is used other tools together, it will improve not only the person but also the overall image quality.

In particular, this is expected to be useful for creators who frequently expose their faces such as V-log and beauty YouTube creators. This is taken with an existing camera, so the naturalness remains the same, and when Face Enhancement is applied after shooting, your face in the screen improves. In addition, the overall picture quality becomes clearer, which can lead to the same effect as the purchase of expensive, high-performance camera equipment.

‘kokoon.cloud,’ a variety of video upscaling services that capture the taste of creators

Upscaling Video Online “kokoon.cloud”

“kokoon.cloud” is an online site where AI upscaling enhance the video image’s quality. In April, the mentioned in this post ‘Remastering’ and ‘Face Enhancement’ functions will be added as well. The upscaling technology currently being offered is also divided into “DeepField-SR1,” which focuses on improving natural image quality and “DeepField-SR2,” which focuses on the removal of noise and emphasis on texture, which can be adapted to the concept of the image.

Visit kokoon.cloud↓

Moreover, it is also the charm of kokoon.cloud that your own upscaling style is created as another function is used. For example, the upscaling technology “DeepField-SR2” can be applied after the aforementioned “Face Enhancement” technology is first applied, or after the “Remastering” technology is applied, video content that revives the character of memories can be created.

Because each technology of kokoon.cloud is excellent, it is possible to receive a satisfactory result even if it is used alone, and the technology can be properly mixed and used according to the direction of the desired improvement. What would it be like when kokoon.cloud’s upscaling service is applied when your own unique content is created in a lot of content?

Golden Age of virtual human, is it a momentary curiosity or will it be an co-existence?

Rosy, a virtual human (Source: Rosy Instagram)

‘Rosy’, a virtual human changed the flow of the industry by giving tremendous ripple effect to advertisement industry. A ‘virtual human’ created by ‘real people’ not only overtakes advertisement industry such as finance, cosmetics and fashion but also shows performance like top stars ahead of debut as a singer. In addition, various virtual humans appeared in row after the appearance of Rosy and we are getting familiar with living together.

When a virtual human appeared at first, there were public opinions that the current virtual human will also be ended with a short curiosity while referring to ‘Adam’, a cyber-singer who failed in the past. However, if you look at what the current virtual human does, she is attracting more attention as she is active in various fields including advertisement industry, entertainment, and influencers. Then, does it mean that virtual humans become an existence who will continue to be with us in the future?

Virtual human, who is ‘celebrity’ contract with entertainment companies.

Yuah, a virtual human (Source: Smile Gate)

Recently, we often hear a news of signing contract between big entertainment company and virtual human. YG Entertainment announced that it has signed an exclusive contract with ‘Han Yuah’, a virtual human of Smile Gate, and ‘Rina’, a virtual human of Netmarble company, had signed an exclusive contract with Sublime, an entertainment company which actor ‘Song Gangho’ and ‘Rain’ belong. Besides, various entertainment companies are announcing their plans for virtual human such as developing their own virtual humans.

As such, virtual humans are extending their field of activity to entertainment business starting as influencers and showing various performances such as singing, dancing, and acting. Since they were characterized by appearance most people prefer, they quickly inform their existence through SNS to get popularity from the public and were able to begin their activities in entertainment industry.

When did virtual humans become popular?

Adam, a cyber singer (Source: Adam Soft)

The beginning of a virtual human in Korea is ‘Adam’, a cyber singer. He made his debut in 1998 and was popular for a while. But his popularity wasn’t continued due to the limitation of 3D graphics at that time, and Adam remained as a failed case. The current virtual humans have overcome the precedent of failure and appeared again has emerged in different situation as before, so many people think it will succeed rather than fail.

The popularity of virtual humans underlies the popularity of Metaverse. Metaverse which combines ‘meta’ meaning virtual and ‘verse’ meaning universe can project our own view of the world into the virtual reality, so it is evaluated as the most promising business. Thanks to contactless culture has become popular due to the pandemic, the public get familiar with virtual world and media also reduced the sense of rejection against virtual humans. In addition, virtual humans are showing activities in various media such as YouTube, Instagram, etc. this made the public accept them naturally.

Evolving AI Human market, possibility of its development is almost limitless!

Another reason the virtual human could grow is graphics. As VR has expanded, the level of graphic has significantly improved than before, so we could realize virtual humans with natural figures just like a real person by adding AI to portraits or videos. As a result, very natural visual that is beyond so called ‘uncanny valley (When human see non-human beings, they feel rather uncomfortable if the level of resemblance is too high)’ was able to create.

Concert of Leslie Cheung that is restored by AI technology (Source: Hong Kong South China Morning Post)

Recently, virtual human based on AI technology can revive the figure in the past. As YouTube became very popular, videos in the past are also actively uploaded. Among them, videos of memory who had already passed away sometimes appeared. However, there were taken too long ago, the quality was not good. However, it was announced that concert tour video of Leslie Cheung in 2000s has restored with 4K quality video.

The concert tour video of Leslie Cheung which is restored by one IT company in China record recording over 10 million views in short time. It said that the background of being able to restore the video was AI. Since AI generated pixel, the figure of Leslie Cheung who was under dark concert hall could be a clear video because AI restored more detailed in the video.

‘kokoon cloud’, a website of improving video quality, is providing a service to improve the quality of videos by using self-developed AI. It allows to enhance video quality by two options such as ‘DeepField-SR1’ that generates the pixel by focusing on subject in the videos and ‘DeepField-SR2’ that erases noise and emphasizes texture. In addition, kokoon.cloud preparing services of ‘Face-Enhancement’ that is optimized for correcting figure videos and ‘Remastering’ that can restore old videos. Therefore, it is expected to be very helpful tools to create figure-related video contents.

‘Virtual Business’ including virtual humans are deeply related with AI. It creates virtual human by graphics and AI, or it gives an opportunity to meet celebrities in our memories who already passed away. In addition, it is not an exaggeration to say that AI involves in all parts of virtual world such as creating characters by compounding AI and graphics, so I am expected to be emerged another virtual world and the role of AI.

Same video quality, but the reduction incapacity

Video content is growing so fast that videos occupy more than 80% of the internet traffic. The mass is actively producing and watching video content. However, the growth of the video content market has some side effects on data. It has become difficult for both the creators that produce videos and video platforms, to handle the enormous capacity of video content.

Then what will be the solution to solve this problem? That is by reducing video data. But reducing video data leads to another problem which is the loss of video quality. Therefore, we need a technology that maintains video quality but reduces capacity.

The reason for the sudden loss of YouTube video quality.

출처: Chubo – my masterpiece / Shutterstock.com

When using YouTube recently, you can notice a big difference. In the past, when you played the video, the video was played in HD quality, but now they are played in lower quality, making us change it back up to HD quality. YouTube supports up to 4K videos, but why does it play it in low-quality videos in the beginning.

The problem is because of the ‘Data issue’. The biggest video platform YouTube tried to reduce the usage, of the sudden increase of the data of video contents by, playing low-quality videos rather than high-quality videos in default. This is an example that the data issue is a serious problem affecting our daily lives.

In addition, the use of smart TVs is also demanding more use of high-quality videos. People using smart TVs enjoy contents, such as YouTube and Netflix, which need high-quality videos. Low-quality videos are more disturbing on a larger screen than on mobile, making the importance of video quality bigger. This makes it impossible for us to give up either the capacity or quality of videos.

‘PQ-optimizer’ is a cognitive image quality optimization technology that save the data rate while maintaining the quality of the video

‘PQ-optimizer’ is a cognitive image quality optimization technology that save the data rate while maintaining the quality of the video. Using self-developed AI reduces data by deleting unnecessary pixels, making it possible to reduce data without the loss of quality.

optimized by PQ-Optimizer

For example, when the computer we use slows down, the first thing that comes to mind is ‘data’. By erasing data, we can make our computers work faster. So we naturally check the usage of data and erase unused documents to maintain data capacity. ‘PQ-optimizer’ also follows this principle allowing users to enjoy the video with lower capacity, by erasing unnecessary pixels in the video.

One thing to notice is that the name of this technology is “cognitive image quality”. In other words, the image quality isn’t the same. But it looks the same in the human eye. Based on the VMAF score released by NETFLIX, when checking the difference from the original video the video that was applied with PQ-optimizer had scored similar. It even had a higher image quality score when using the same amount of data.

The difference ‘PQ-optimizer’ can bring in daily life.

You can see that by using ‘PQ-optimizer’ we can improve coding efficiency to reduce the capacity of videos but maintain quality, allowing us to produce more video content. Thanks to this technology we can produce and enjoy video content freely. Since it uses less data, it can also transmit high-quality videos in unstable network conditions which can be useful in live streaming.

In addition, the PQ-optimizer is a pre-processing technology that can be applied before the encoding system and because it can be applied to all codecs, it does not need the expansion of any new codecs. It can also be encoded better in the same given data and has better encoding efficiency when the quality is the same.

New creators are emerging day by day and they are producing video content actively. The mass also enjoys a lot of video content. As the amount of content is increasing tremendously, we need to prepare for the usage of data. The ‘PQ-optimizer’ introduced today will be very useful in the upcoming data issue.

*Please contact the email below if you want more information about the ‘PQ-optimizer’ introduced in this post.

Contact: info@blue-dot.io

Upscaling, the best solution for retouching videos.

Facebook, Instagram, and Youtube… SNS have become deeply inside our life. It is because People can record their daily life easily compared to before when we should carry camera to record our memories and stories. As there are various people using SNS, various contents are being upload. However, there is one thing in common that they go through the “retouching” stage.

After taking pictures, retouching stage has become a necessity, not an option. Therefore, various camera applications for retouching photo are being developed and increased popularity steadily. There are a lot of camera applications, but video retouching applications are relatively rare. So I propose that Instead of applications, you can use ‘upscaling’ as a solution that easily retouching videos.

SNS is addicted to video, so I record everything in videos!

SNS tends to prefer videos. For example, when we use Instagram, most of the first appearance popular post when we searching contents is video and they launched ‘reeels’, a platform dedicated to video contents. In particular, ‘Reels’ has become a popular video platform in a short time because they have many similar aspects to video SNS ‘TikTok’. Besides, as YouTube grow rapidly and the number of creators who show their daily life through V-log has increased, more and more people upload video to SNS as a means of recording their memories.

However, V-log without retouching is likely to be ignored in huge SNS’s contents world, because most creators improve the quality by retouching color and skin tone of the video. Therefore, videos without retouching, the ‘raw’ videos seem to relatively low-quality contents. In order to solve this problem, many creators tend to use camera applications that have been retouched from the beginning, and sometimes purchased high-performance cameras.

Both camera application and high-performance camera can enhance video’s quality, but they have disadvantages clearly. First, camera application can retouch skin-tone and highlight beauty of face, but not guaranteed the video’s quality. In addition, there is a risk that decrease quality of the video, shot by camera application, when you watch on a bigger display such as tablet PC or desktop. High-performance camera can make clear video quality on any display, but it is very expensive.

Natural retouching and ease, the best retouching video solution ‘upscaling’.

Upscaling is the best video retouching solution which is able to retouch video naturally while compensate the defects. Upscaling is a technology that enhance video quality, and it can be used freely depending on various purpose of use. For example, even if the resolution of the video is low, upscaling helps to watch clearly on a large display such as TV or if you want retouching whole video’s color to brighter, upscaling could be very useful.

Recently, upscaling based on AI is getting attention. Previously, small video should be blur because of pixels are simply copied and pasted average value of the nearest, but if you use AI to upscale video, deep-learning model expects and generates pixels so it could be retouched naturally and clearly. For example, I upscale my face video to 4K then my skin-tone will be brighter and clearer and accentuate the contour of face, so it makes my strengths stand out.

Another reason why upscaling is the best video retouching solution is ease. Especially, it can use not only software but also online, kokoon.cloud is video upscaling online service to upscale your video up to 4K. When you upload a video, it converts the video to desired size, AI generates pixels to natural upscaling. In addition, the conversion time is very fast to upscale video with the best quality, and you can use everywhere web-capable place.

Upscaling is an optimal video retouching solution that is very efficient in terms of time and price for natural retouching. In strong competition of various video contents, many creators try to make better contents to attract audiences. Wouldn’t it be possible to become winner in this competition if you enhance the quality of videos easily and quickly through upscaling? Try upscaling, easy video retouching solution everyone can use.

Online video upscaling ‘kokoon.cloud’ renovated open.


On 24th Mar, online video upscaling based on cloud service, ‘kokoon.cloud’ renovated open. Kokoon.cloud is getting famous at home and abroad because they serve fast and easy online video upscaling, and it can upscale up to 4K and great quality. We decide to renovate website for better service to more users.

Through this renovation, we focused on maintaining ease of use as before and adding new tools so that users upscale with higher quality. First, separate workspace and add preview tool, especially we introduce new upscaling version that added new AI models. Users can get upscaling results tailored to their favor.

1. Make workspace for focusing on upscaling

kokoon.cloud workspace

Kokoon.cloud separate main page and workspace to users can upscale their video more easily. In addition, users can set details like resolution, video format, quality, etc. to taste, it calls ‘option’. You can make various options on a video to different results.

Even though workspace is separated and provides more sophisticated upscaling through setting options, kokoon.cloud maintains the ease of use. Upload your video on workspace and set option on your way, then click ‘PROGRESS CONVERTING’ button to upscale. If you click the original video, only display the video’s result, so users can see which one is the upscaling video without being confused.

2. ‘Preview’, you can compare B/A before download.

kokoon.cloud preview

‘Preview’ is new tool in workspace to see upscaled result before download. Click the preview in option window and compare the original video and upscaled result based on the options you set. You can drag the middle bar to move freely to check whether the areas that need to be emphasized or there are any problems.

If you want to compare before and after more precisely, then adjust screen magnification from 100% to 400%, or scale to fit. Especially, kokoon.cloud have two versions of upscaling that apply different AI models, so user can get a better upscaling result by comparing which version goes best with the video.

3. Two versions of upscaling to satisfy user’s style.

Kokoon.cloud added new features for new website, the most notable part is ‘Deepfield-SR2’, another version of upscaling. Kokoon.cloud’s upscaling is based on BLUEDOT’s ‘Deepfield-SR’, recently BLUEDOT developed new AI model to meet user’s needs. It applied to ‘Deepfield-SR2’, and both can use in kokoon.cloud.

‘Deepfield-SR1’is same as before, AI generates pixels focusing on objects, so it enhances video naturally and quickly. The other version of upscaling, ‘Deepfield-SR2’, is focused on erase noise. it deletes only noise and emphasize textures, so it upscale more moderate than ‘Deepfield-SR1’. Two versions of upscaling have different characters, you can choose on your style.

4. The first impression of kokoon.cloud, main page.

kokoon.cloud how to use?

The main page of kokoon.cloud introduce what kind of work we do to guests. ‘How to use?’ is a tutorial video showing how to use workspace, and users send any opinions like improvements or concerns using ‘Talk to Sales’. Above this, it introduces kokoon.cloud’s communities like YouTube, Facebook, and Blog.

kokoon.cloud is planning to make various attempts from this renovation. We provide other useful video solutions except online video upscaling and better services by communicating with kokoon.cloud users through newsletters and SNS communities. All of this is thanks to steady increase kokoon.cloud’s users and the continued positive evaluations of our upscaling quality.

Why YouTube supports 4K videos?

More than 50 billion worldwide creators upload about 24,000 hours of contents in an hour, and more than 2 trillion users watch these huge number of contents. The value is recorded by YouTube, the word’s largest video platform.

Sometimes YouTube to be a standard for video contents. What is the trendiest contents? Who is the rising creator today? When I curious those information, I always search in YouTube. As such, YouTube is leading video contents market with steady changing and innovating. And supporting 4K videos is one of the various and innovative changes in YouTube. 4K has more pixels than HD about 4 times, so it express video definition in high quality. Then I curious a question, why the video trend leader YouTube supports 4K videos?

4K, the symbol of high-definition videos.

4K means there are about 4,000 pixels in the video horizontal. It also used to refering Ultra-high definition resolution and called 4K UHD(Ultra HD). 4K video’s size is 3840×2160, which is 3 times bigger than HD, and the quality is outstanding because it has more pixels than Full HD 4 times. When 4K appeared, 4096×2160 and 3840×2160 were used together but CEA(Consumer Electronics Association) set the size 3840×2160 in 2012, which is the current standard.

In 2012, South Korea transmitted 4K video for test broadcasting and start the era of 4K from while live broadcast 4K UHD some matches in Brazil world-cup 2014. Since then, 4K display market has grown rapidly because 4K contents has increased, now we watch 4K videos in daily life like desktop, laptop etc.

YouTube attempted to build a platform to watch 4K videos in 2015 for preparation the popularization of 4K videos. First, they test platform operation by playing 6 4K videos, and then provide 4K resolution function. YouTube provides SD, HD, 4K resolutions, and the highest resolution user can watch is 4K. In conclusion, supporting 4K on YouTube, the largest video contents platform, means that 4K has become the standard of high-quality video in the world.

Satisfied the public addicted to watching!

Not only YouTube, OTT platforms such as Netflix also play 4K contents because the vido trend have been changed. From 4K display is commercialized quickly, people watch videos on various displays like desktop, TV and they want to keep high quality on any displays.

Recently, Google Chromecast is very popular in South Korea. Even before release, many people order it online from overseas. In addition, the new version of TVs support video platforms such as YouTube and Netflix, Google Chromecast is gaining popularity more and more.

This means that people who watch videos on mobile to TV display have increased. Video content area has become huge thanks to people watch videos increased rapidly and videos become diverse for people who want to enjoy spare time. Therefore, people are inconvenienced when they watch video content on a small mobile screen for a long time, so they prefer to bigger display to watch video more comfortably.

Is the successful 4K video at stake?

However, I am worried about the future of 4K video because the number of contents provided in 4K is insufficient to satisfied public who are looking forward to high-quality video content. Besides, some defects that it is expensive to produce 4K contents and if the popular content you previously owned is not 4K, we cannot meet the demand to watch it in high quality.

Upscaling is the most effective solution to solve the problem. Upscaling can convert the video finished shot to 4K, it means that you can enhance HD video 3 times to 4K thanks to upscaling. It is much cheaper than shot 4K video through high-performance camera and even convert the video you already have. Video creators can use upscaling very usefully to make high quality contents. Kokoon.cloud is fast and easy video upscaling online service, anyone can upscale video through website for free.

Visit to kokoon.cloud ↓

Video resolution will be important factor to evaluate and YouTube which is leading the trend of video contents are realized it, so they play 4K video to public. I think the decision have affected users in YouTube to increase.

How to keep high quality of video on any displays?

With the spread of smart home appliances linked to mobile to the public, more people enjoy video contents such as YouTube and Netflix on larger displays. It is more comfortable and realistic when you play the video on a larger screen than mobile.

However, if you watch mobile-size video on a big screen more than 50 inches arise a big problem. The quality of video gets blurry when you increase the video size. This is a natural result, but people can’t stand this inconvenience. We need to solve this problem and ‘upscaling’, keeping high quality of video on any displays, is the most efficient solution among any technologies.

Upscaling is a technology that global companies such as NVIDIA, AMD, and Microsoft are paying attention to.

upscaled by kokoon.cloud

Upscaling is a technology that enhances a video to keep quality in detail while increasing video resolution. Today the scope of upscaling is gradually expanding, because the resolution of displays is increasing but the video contents cannot keep up with the pace of screen size. 4K display has already been supplied and now we even watch 8K displays. However, the number of 4K videos are not enough to play and while old videos before HD resolution are popular again, the application of upscaling technology is gradually expanding.

Source: NVIDIA

In game industry, upscaling is already being used to improve gaming user experience on PC. NVIDA, global IT company, developed DLSS and get ahead of game upscaling part, and AMD also participate into the upscaling technology field by FSR. Recently, even Microsoft has announced that they are researching upscaling software using AI. As a result, “Upscaling” become the most noticeable keyword in the industry.

Sound like a dream 8K display, but no contents?

Source: Samsung

There being more reason why upscaling is on the rise, that is a correlation between the development of display technology and the number of video contents. The sales of 8K TV had been steadily increasing since they appeared, but it is slowing down lately. Because people are looking forward to watching on 8K TV very high resolution video which is four times higher than 4K TV, but there are few contents to satisfy 8K TV’s resolution. None of broadcasting stations provide 8K video yet, and OTT platforms such as YouTube, Netflix, and Apple TV also service up to 4K videos.

In current contents creating environment, it is difficult to produce UHD videos such as 4K or 8K. First, cameras equipment which is capable of shooting 4K video is expensive. Although UHD camera is emerging, but video quality is often not good depending on the shooting environment. In addition, it takes long time to produce, so it is very inefficient in terms of cost and time. For these reasons, it is being researched to produce 4K contents by converting to 4K using upscaling to enhance video resolution after producing content at 2K resolution.

Is upscaling professional? NO! everyone can use it.

Even though the scope of using upscaling is expanding and it is a rising technology, but it is still unfamiliar term to public. Is upscaling a specialized technique that can only be used by specialists? No! Upscaling is a very simple technique that anyone can use.

Today, there are more than 2 billion worldwide people using YouTube, and the number of creators is about 50 million, So, in this era where anyone can create and enjoy video content, and the number of users is increasing, video quality improvement technology that enhances the contents is necessary for everyone. For this reason, upscaling technology providers are providing services through software and online so that anyone can experience upscaling easily and simply.

Kokoon.cloud is easy and fast online video upscaling through cloud and we serve online upscaling for free. Just upload your video and start converting, you can make high resolution video without losing quality. We serve our upscaling technology for free so that anyone can experience.


As the video contents market grows, resolution has become a factor in determining the overall quality of content. In addition, the number of contents is increasing, and as display technology is advanced, public users consume contents on more than one equipment. Therefore, upscaling that can achieve high resolution on any display is no longer an option but a necessity.

AI technology, How far will it develop?

Today, we ask machines questions that may be asked to people such as ‘Tell me the title and singer of the song playing right now’, ‘What’s the most popular movie in cinema these days?’, ‘How is the weather today?’. However, Now is the era where no one finds it weird, this is thanks to ‘AI (Artificial Intelligence)’ that permeated our lives as a whole.

This technology called ‘Artificial Intelligence’ or ‘AI’ is a computer technology that artificially implements human learning ability, inferencing ability, and perception ability. Because AI can apply what humans have to do to machines, it has many advantages, such as a wide range of applications, reducing labor costs, and replacing some of people’s roles in the current situation where the population is increasingly scarce.

The life with ‘AI’ that now we are accustomed


The opportunity to imprint unfamiliar Artificial Intelligence on the public was the Go match between AlphaGo and Lee Se-dol held in 2016. As the match between AlphaGo, an Artificial Intelligence Go program developed by Google’s DeepMind, and Lee Se-dol, a human who boasts the highest Go skills, drew public attention, the public could see what role the Artificial Intelligence plays. Since then, various services with Artificial Intelligence have been created, and the life with AI has begun.

Representatively, there is an Artificial Intelligence speaker. The artificial intelligence built into the speaker acts as a secretary, enabling conversation like humans, and providing information. It was a rather unfamiliar product, but it boasts quite a lot of popularity and has become a common item in our daily life. AI is preferred by office workers who live alone because of its various function such as listening music, weather prediction and alarm, and it plays many roles such as being a conversation partner for the elder who lives alone in a situation where face-to-face contact has become difficult due to COVID-19.

Autonomous driving is also a representative industry in the field of applying Artificial Intelligence. Autonomous driving reduces the energy consumption of drivers by helping cars recognize traffic conditions itself and drive safely, which shows how much AI contributes to human life and proves the high value of technology. Actually, Hyundai Motor Company is making generous investments and showing continued interest by selecting AI technologies such as autonomous driving and Urban Air Transportation (UAM) as key technologies that will be responsible for the future.

The more upgraded world of Graphics due to AI

Also, another field that has drew attention due to the application of AI is ‘Graphic’. Using Artificial Intelligence technology, viewers are amazed by real graphics because it can realize more realistic graphics by reducing the difference between the screen and reality. Additionally, A new area has been also created as the level of graphics have been upgraded surpassingly due to application of AI.

Virtual Influencer ‘ROZY’ (Shinhan Bank)

‘Virtual influencer’, which was created as a virtual human being by applying AI technology to the face of a real person, is now the best model. Virtual influencer does not cost labor, can be controlled, and has no time and space limit, so it can be used in many places at the same time. As a result, it is very efficient in terms of time and cost, so various brands are showing interest in marketing content in collaboration with Virtual influencer. The popularity of Virtual influencer, which appears in advertisements of almost all brands, can be attributed to the fact that Deepfake technology using AI has reduced the difference between reality and virtuality.

Upscaling, a technology that improves video quality, also brought about further upgraded results thanks to AI. The existing upscaling technology, which enlarges the video by copying the average value of pixels, often can increase the resolution of the video, but the quality becomes blurred. However, When Artificial Intelligence technology is applied, the trained AI inferences and generates pixels that best match the video at high resolution, so the video not only enlarges but also maintains clear image and high quality.

KOKOON Cloud’s upscaling technology also applied AI. Its proprietary Artificial Intelligence trained from numerous video data generates the most suitable pixel, enabling the best quality improvement that is natural overall and does not lose details even if looking closely. Even if you enlarge the 640x360p SD video to 3840x2160p 4K, you can see videos with improved quality without breaking off.

On the other hand, there are voices of concern about now that AI technology is deeply entering our life. It hasn’t been long since AI was involved in our life, and various trials and errors may occur because we are currently in the stage of establishing a direction to life with AI. But the clear thing is that thanks to Artificial Intelligence, the quality of our life will be better and we live a more wealthy and convenient life.

AI 기술의 발전, 어디까지 갈 수 있을까?

‘지금 들리는 노래의 제목과 가수를 알려줘’, ‘요즘 영화관에서 가장 인기있는 영화는 뭐야?’, ‘오늘 날씨 어때?’ 등 사람과의 대화에서 이루어질 법한 질문을 기계에 합니다. 하지만 그 누구도 이상하게 여기지 않는 시대가 되었는데요, 현재 우리 생활을 전반적으로 지배하고 있는 ‘AI(Artificial Intelligence)’ 덕분입니다.

‘인공지능’ 또는 ‘AI’로 불리고 있는 이 기술은 인간의 학습능력과 추론능력, 지각능력을 인공적으로 구현하고자 하는 컴퓨터 기술입니다. AI는 인간이 해야 하는 일을 기계에 적용시킬 수 있기 때문에 활용 영역이 광범위하고 인건비를 줄일 수 있으며, 인구가 부족해지고 있는 현 상황에서 사람의 역할을 일부 대신할 수 있다는 점 등 많은 장점을 가지고 있습니다.

이제는 익숙해진 ‘AI’와 더불어 사는 삶

출처: 구글

생소했던 인공지능을 대중들에게 각인시킬 수 있었던 계기는 2016년 개최된 ‘알파고vs이세돌’의 바둑 대결이었습니다. 구글의 딥마인드에서 개발한 인공지능 바둑 프로그램인 ‘알파고’와 가장 높은 바둑 실력을 자랑하는 인간 ‘이세돌’의 대결이 세간의 주목을 받으면서 대중들은 인공지능이 어떤 역할을 하는지에 대해 알 수 있었죠. 이후로 인공지능을 적용한 다양한 서비스가 탄생하며 AI와 더불어 사는 생활이 시작되었습니다.

대표적으로 ‘인공지능 스피커’가 있습니다. 스피커에 내장되어 있는 인공지능이 비서의 역할을 하면서 인간처럼 대화도 가능하고, 정보를 제공해주기도 합니다. 다소 생소했던 제품이지만 제법 많은 인기를 자랑하며 일상에서 흔하게 볼 수 있는 아이템이 되었는데요. 음악 감상부터 날씨를 알려주기도 하고, 알람 기능까지 갖추고 있어 혼자 자취하는 직장인들에게 선호되고 있으며 코로나19로 대면 접촉이 어려워진 상황에서 집에 홀로 남겨진 어르신들의 대화 상대가 되어 주기도 하는 등 많은 역할을 해주고 있습니다.

자율주행 역시 인공지능을 적용한 분야의 대표주자입니다. 자율주행은 자동차가 스스로 교통상황을 인지하고 안전하게 주행하도록 이끌어줘 운전자의 에너지 소모를 줄여주는데요, 이는 인간의 삶에 AI가 얼만큼 기여하는지를 보여주며 기술이 보유하고 있는 높은 가치를 증명해주기도 합니다. 실제로 현대자동차는 자율주행과 도심항공교통(UAM) 등 AI기술을 미래를 책임질 핵심 기술로 선정하면서 아낌없는 투자와 지속적인 관심을 보이고 있죠.

AI 덕분에 한층 업그레이드된 그래픽의 세계

그리고 AI의 적용으로 주목받게 된 또 하나의 분야는 ‘그래픽’입니다. 인공지능 기술을 활용하면 화면과 실제의 차이를 줄여 더욱 실감나는 그래픽을 구현할 수 있기 때문에 보는 사람들은 실제와 같은 그래픽에 감탄을 하게 되죠. 게다가 AI 덕분에 뛰어난 그래픽 수준을 갖게 되면서 새로운 영역이 탄생하기도 했습니다.

버추얼 인플루언서 ‘로지’ (출처: 신한은행)

실제 사람의 얼굴에 AI 기술을 적용해 가상의 인간으로 탄생시킨 ‘버추얼 인플루언서’는 지금 가장 잘나가는 광고 모델입니다. 버추얼 인플루언서는 인건비가 들지 않고, 통제가 가능하며 시공간의 한계가 없어 동시에 여러 곳에서 사용할 수 있습니다. 때문에 시간과 비용 면에서 매우 효율적이어서 다양한 브랜드는 버추얼 인플루언서와 협업한 마케팅 콘텐츠에 관심을 보이고 있는데요. 거의 모든 브랜드의 광고에 등장하고 있는 버추얼 인플루언서의 인기는 AI를 활용한 딥페이크 기술이 현실과 가상의 차이를 줄여준 덕분이라고 할 수 있습니다.

영상의 화질을 향상시켜주는 기술인 업스케일링도 AI로 인해 더욱 업그레이드된 결과를 가져올 수 있게 되었습니다. 화소의 평균값을 복사해 영상을 확대한 기존의 업스케일링 방식은 영상의 크기는 커지지만 대신 화질은 흐릿해지곤 하는데, 여기에 인공지능 기술을 적용하게 되면 학습된 AI가 스스로 영상에 가장 어울리는 화소를 찾아 생성해주기 때문에 영상의 크기가 확대될 뿐 아니라 화질까지 또렷하게 유지해줍니다.

코쿤클라우드의 업스케일링 기술 역시 AI를 적용했습니다. 수많은 비디오 데이터를 통해 학습한 자체 개발 인공지능이 가장 어울리는 화소를 생성해줘 전체적으로는 자연스러우면서 자세히 살펴봐도 디테일을 잃지 않는 최상의 화질 향상이 가능합니다. 640x360p의 SD급 영상을 3840x2160p의 4K급으로 영상을 확대하게 되더라도 깨짐 현상 없이 자연스럽고 또렷하게 화질이 개선된 영상을 만나볼 수 있습니다.

코쿤클라우드 바로가기↓

한편에서는 AI기술이 삶에 깊숙이 들어오는 지금을 걱정하는 목소리가 들리기도 합니다. 하지만 아직 우리의 생활에 AI가 관여하게 된 시간이 얼마 되지 않았고, 현재는 AI와 함께 살아가는 방향을 정립해가는 단계이기 때문에 다양한 시행착오가 발생하게 될 수도 있습니다. 하지만 분명한 사실은 인공지능 덕분에 삶의 질은 향상될 것이며 우리는 더욱 윤택하고 편안한 생활을 살아갈 수 있을 것입니다.