How does the virtual human live broadcast of station B work?

The virtual human live broadcast of station B is mainly realized through virtual anchor technology. Specifically, the virtual anchors of station B are produced using virtual reality technology and artificial intelligence technology, including 3D modeling, animation production, speech synthesis, facial expressions, etc. The virtual anchor can realize real-time motion capture and facial expression capture through the motion capture system, so that the audience can see the real movements and expressions of the virtual anchor in the live broadcast.

The virtual anchors of Bilibili usually have their own unique images, personalities and characteristics, such as Hatsune Miku, Googo-chan, Kodori, etc. Virtual anchors can interact with viewers by conducting live game streaming, singing, chatting, and other activities.

How to make a virtual streamer? How to use virtual anchors?

Virtual anchors are currently relatively hot in the field of Internet short ** and live broadcast. Most Internet celebrities and two-dimensional anchors have considerable traffic. Some Internet celebrities have tens of millions of fans, and they have a good income from live streaming and advertising and monetization.

But many people want to be virtual anchors but don't know where to start, how to make virtual anchors, how to apply virtual anchors, how to broadcast live broadcasts, etc., not much to say, let's talk about how to do virtual anchors, and some problems that will be encountered.

1: How to make a virtual anchor?

To make a virtual anchor, you first need to have a virtual image that you want to do, and then you can choose different ways according to your budget.

2D Virtual Anchor Reference.

For example, if you want to be a 2D anchor, then you can go to station B to find some free software to live broadcast, the software can generally choose your own virtual image, and then you can use the computer camera to drive the virtual anchor's expression in real time to live broadcast.

3D Virtual Anchor Reference.

The production of three-dimensional virtual anchors is more complicated, first of all, you need to design the image, and then carry out 3D image modeling, after the modeling is completed, you need to make the bones of the expression and the body, render the AR environment, and finally need to bind to the software or motion capture equipment for real-time drive, if you want to drive the whole body and the expression in real time, then the commonly used is to use the motion capture equipment, and the motion capture equipment also needs a lot of investment, the motion capture equipment on the market generally ranges from tens of thousands to hundreds of thousands, If the requirements are high, it is not recommended to consider devices below 50,000. There are a lot of pits in the motion capture equipment, and you can only understand it after stepping on it.

Virtual human live streaming looks quite complicated, but it's actually very simple. I see that many people on the Internet spend hundreds of thousands to learn this technology, but in fact, there is no need for this at all.

Virtual human live tutorial.

How to adjust the virtual character of the live broadcast.

You need to make an image with 3D modeling. In China, virtual anchors are commonly known as virtual UP masters, virtual uploaders (VUPs). Outside of China, virtual streamers are known as Virtual Youtubers (VTubers) due to their widespread presence on YouTube.

Introduction:

Looks like you want to be a VUP.

I advise you to give up.,Originally, the overall popularity (especially the strength of the money) has always been VUP" background and image for facerig

The barrage uses the barrage Ji.

Here's how a virtual anchor works out:

1. Customize and design one for you through the artist, and then import it into the virtual live broadcast software, such as the VTMINI of Super Dimension, support the import of local characters, or you can use the virtual live broadcast software with its own virtual image, in VTMINI, there are also many virtual images to choose from, here you can choose your favorite virtual image, and in the action resources, you can choose a variety of character actions.

2. For hardware equipment, you must first have a computer with sufficient configuration, then you need a camera to be responsible for the facial capture part, and finally headphones and microphones. Headsets are no problem, but if possible, it is best to have equipment such as headphones + independent microphones + mixers. In the part of the software, VTmini also has motion capture and face capture functions, which can capture the facial expressions and body movements of the person in real time.

How does a virtual anchor make a debut?

1. Debut as an independent person through spontaneous search for resources and planning. This way is close to entrepreneurship, not restricted by the company, and entering the market competition as a similar anchor, which is actually similar to the anchors on platforms such as Douyu and Huya.

2. Participate in virtual planning and become a corporate vup. For example, the domestic virtual research company, or the Sino-Japanese cooperation virtualreal, are both corporate potential VUP cultivation plans, such VUP can obtain the support of the company, and then enter the market competition.

The technical threshold required by virtual anchors is high, and they generally need two things, one is software, and the other is hardware.

On the hardware side, at least one computer is needed to drive a series of virtual live broadcasts.

The software can refer to a company on the market, one-click import** can generate a character model, people, goods and yards are all supported, whether it is live broadcast or live broadcast with goods, it can be perfectly covered.

To be a virtual anchor, the hardware with the lowest threshold is a smartphone, and because its cost is low enough, the level of detail of character models and motion capture is relatively limited.

Virtual live streaming scenarios.

The live broadcast plan can be set up in advance in the virtual live broadcast tool, the so-called live broadcast plan is the role matching scene during the live broadcast, the lens, the sound capture parameters of each side, according to your needs, customize multiple sets of parameters, which live broadcast plan you need, you can switch it with one click, which is very convenient.

Start settings. We can choose a good mode according to the category of the model (3D model or live2D model).

Here we use 3D mode login as an example.

In the process of customizing the model, you need to convert the model format to PMX, FBX, and VRM formats.

Set expressions and actions.

VUP can set expression shortcut keys, VUP expression channel includes: automatic blink channel, face capture channel, sound capture channel, automatic camera viewing channel, please map the required expression channel according to your needs.

The face capture VUP supports face capture with both ordinary cameras and mobile phone cameras, and you need to calibrate the face capture before the live broadcast starts. Before face capture, we need to select the camera or mobile phone for the live broadcast and enable face capture. [VUP surface capture parameter setting].

MoCap VUP supports most of the mocap hardware on the market. Tutorial on how to use each mocap device.

Sound capture. Mobile phone virtual camera software vface

Install the software and set it up.

Set the resolution and image quality.

Next, you need to set the resolution and quality of the image according to the actual situation of your computer, this indicator is different for each computer, and I can't give a specific value here. If you are not sure about the configuration of the specific computer, you can directly check the computer properties.

Set the green screen.

It is applied to live streaming platforms.

Here is an example of a live broadcast station, you need to identify the camera as a facerig, and then adjust the character value to green, so that the virtual portrait will replace the real avatar during the live broadcast.

Set up a virtual background.

The final results are presented.

After choosing the background, the picture will be displayed as shown in the figure below, and you can also adjust the character lens through the ALT + mouse wheel. The same is true for the live broadcast on the mobile phone, as long as you find the right software, you can set yourself as a virtual character for live broadcasting if the corresponding settings are intact.

Speaking of live broadcasts, many friends are familiar with them, and live broadcasts can often be seen on major social platforms, generally speaking, live broadcasts are divided into many kinds, some singing and dancing, and some live broadcasts selling goods.

, all kinds of live broadcasts emerge in an endless stream, attracting many people**. Normally, most live broadcasts are real people, but with the continuous development of technology, there are many ways of live broadcast of avatars, which are also slowly emerging, at station B.

You can see that there are a lot of avatar live broadcasts. Virtual live streaming is not only an alternative to live streaming, but it also has a lot of unique advantages.

In daily life, many live anchors have a certain time limit when they broadcast live, and it is difficult to broadcast live 24 hours a day. But virtual anchors are different, they can be said to be tireless, even if it is a 24-hour loop live broadcast. Not only that, but virtual anchors can also continue to live broadcast in place of live anchors after the live anchor goes off the air.

Although virtual anchors can broadcast live for a long time, their movements and speech skills are still slightly worse than those of real anchors. Basically, some templates are designed in advance through virtual technology and real-time animation technology, and when the virtual anchor broadcasts live, people's questions and demands will be asked according to the pre-set template. In addition, the cost of using avatar live streaming will be lower than that of real anchors, but there are still many disadvantages.

In general, there are pros and cons to virtual image live broadcast, and it is believed that in the future, AI technology will continue to improve, and the problem of interaction between virtual anchors and people will also be solved. At present, there are already some application scenarios for virtual main jujube broadcasting, such as helping live broadcasts to bring goods, or chatting, etc., during live broadcasts, interaction with users can also be timely problems. It is believed that in the future, the combination of real and virtual image live broadcast will definitely do better and better.

I don't think so, the virtual image is not a real person, and it is difficult to give correct answers to the questions raised by netizens, and it cannot replace the live broadcast.

Behind the virtual characters is a real person controlling, which is no different from a live broadcast. This technology can improve the user experience, enhance user stickiness, and choose the favorite character to control the scene.

Virtual character anchors are just like high-tech, they will have a lot of problems from netizens, but the number is limited, and the anchor's tasks will be divided into several sections.

In the narrow sense, the virtual anchor uses the original virtual personality setting and image to carry out activities on **Inspection** and social platforms. Most of the characters appear as 3D models of MMD or Unity, or 2D models made by Live2D, and are voiced by real voice actors, but voice actors are generally not publicly available. **There are many forms, vlogs and live games.

By this measure, Amiyamato can be considered the world's first virtual YouTuber. However, it is only an animated image instead of its own image for submission, and it has a real identity behind it, which can be considered a virtual youtuber in a broad sense.

It is to replace the anchor itself with a virtual anime character.