RelaxKiral
UE5 Highlights, Import Metahuman and Switch to the main Character in UE5 Sample Project Tutorial
updated
In this step-by-step beginner tutorial, I’ll show you how to set up ComfyUI with the Wan-Animate Workflow on a low VRAM machine (12GB+ NVIDIA GPU). This method makes it possible for more people to try AI-powered talking avatars and bring a single image to life!
Whether you’re new to ComfyUI or just looking for an optimized setup for Wan-Animate, this guide covers everything from installation, fixing common errors, to creating smooth animated avatars.
🔗 Download files + full instructions here:
👉 patreon.com/posts/140102635
🖥️ What You’ll Learn in This Tutorial
Install & configure ComfyUI for Wan-Animate
Run the workflow in low VRAM mode (12GB NVIDIA GPU required)
Set up CUDA Toolkit 12.9 and custom nodes
Fix common issues like SageAttention, Triton, and embedded Python errors
Load missing nodes & optional Crytools setup
Create better masks for smoother results
Debug and fine-tune workflow parameters
Export & review your final talking avatar
⏱️ Video Chapters
0:00 Intro
1:27 Wan-Animate Overview
2:29 ComfyUI Setup
5:30 Install CUDA Toolkit 12.9
6:26 ComfyUI Manager (Custom Nodes)
9:02 SageAttention Fix
11:34 Triton for Windows
11:57 Embedded Python Fix
14:14 Start ComfyUI
15:59 Open the Workflow
16:39 Debug + Fix Errors
18:22 Install Missing Custom Nodes
20:33 Workflow Loaded (No Missing Nodes)
20:51 (Optional) Crytools Install
22:12 Download + Place Models
27:49 First Run Test
31:29 Create Better Mask (Point Editor)
34:51 Export Result File
35:17 Check Results
36:22 Workflow Parameters Explained
38:01 Final Result
🚀 Why This Tutorial?
Many people asked how to set up Wan-Animate with ComfyUI after my last post, so I made this beginner-friendly guide. Your feedback and reactions help me create more tutorials like this!
If you found this video useful, don’t forget to:
👍 Like
💬 Comment
🔔 Subscribe for more AI animation, Metahuman or ComfyUI tutorials
📚 Reference Links
🔗 Wan-Animate: humanaigc.github.io/wan-animate
🔗 ComfyUI: github.com/comfyanonymous/ComfyUI
🔗 Example Workflow (WanVideoWrapper): github.com/kijai/ComfyUI-WanVideoWrapper/blob/main/example_workflows/wanvideo_WanAnimate_example_01.json
The future of content creation, VTubing, and digital identity is here, and it's can run on our local PC!
In this video, I'm showcasing the incredible power of bring the ref Avatar to live. I was blown away by the results.
If there is enough good reactions, I might do a tutorial on how to do this in different levels.
The Magic Recipe:
1 Picture: Your favorite anime character, a drawing, or even a stylized photo.
1 Voice Clip: An audio recording of what you want your avatar to say.
1 Motion Source: A short video to provide the head movements and expressions.
Combine them with the local AI, and you get a perfectly lip-synced, expressive, animated character!
This is a massive breakthrough for anyone who wants to create an animated presence online without expensive software or privacy concerns.
Details:
patreon.com/posts/turning-1-into-139397316
tutorial video
youtu.be/N_vzgnFY45c
In this video, I'm putting the new FLUX.1-Kontext model to the test to see just how far we can push it, right on my own computer. This is a game-changer!
We're going beyond simple filters and exploring complex, context-aware edits. Watch as we use simple text prompts to perform incredible transformations while maintaining the original image's composition and style.
🧪 *WHAT WE TESTED IN THIS VIDEO:* 🧪
00:00 - Removing People & Objects Instantly
00:21 - Changing the Entire Environment (from sunny to rainy)
00:30 - Changing Text and keep the style
00:43 - Editing a Character's Pose & Camera Angle
00:55 - Changing Clothing Colors
01:06 - Transforming an Anime Character into a Realistic Person
01:31 - Restore the image color and remove scratches
01:47 - Change the character to face the front and side angle
02:11 - Advanced Test: Editing In-Image Text & Realistic Transform Simultaneously
02:59 - DONE :)
This workflow demonstrates the power of context-aware image generation, where the AI understands what to change and, more importantly, what to leave alone. The ability to edit characters, backgrounds, and even text with this level of precision is incredible.
---
🧠 *THE MODEL BEHIND THE MAGIC:*
This experiment was done using the FLUX.1-Kontext development model. You can learn more about it and try it for yourself here:
huggingface.co/black-forest-labs/FLUX.1-Kontext-dev
What do you think of these results? Let me know in the comments what I should test next! :)
Don't forget to like and subscribe for more AI experiments or Tutorial on the relaxKiral channel!
This video is for learning and educational purposes
PC specification for this video:
Processor AMD Ryzen 9 7950X3D 16-Core Processor 4.20 GHz
Installed RAM 64.0 GB (63.6 GB usable)
RTX 5090 Blackwell
#AIPower #PhotoEditing #ComfyUI #StableDiffusion #FLUX #Kontext #LocalAI #TechDemo #AIArt #GenerativeAI
1. Using ChatGPT to create your image with prompt.
"draw in anime, Whimsical, nostalgic, magic, and everyday life style, keep the background. Keep the ratio of the image."
I used this video screenshot from my channel:
youtu.be/x37uIyt7K2Y?si=yeNq1_0UoYo7flPQ
2. Use audacity or any other voice recorder to record your voice and export it to a .mp3 file.
audacityteam.org/download
3. Using hedra.com with prompt
Upload your image from (1)
and
Upload your voice from (2)
In prompt I used
"man speaking on podcast"
You should change it accordingly
If you like it, please give it a thumbs up :)
This can potentially help on the 3D creation process
It is open source and here is the github link that I used
github.com/kijai/ComfyUI-Hunyuan3DWrapper
Model that I used
huggingface.co/tencent/Hunyuan3D-2mv/tree/main/hunyuan3d-dit-v2-mv
Have fun :)
- IT is LOCAL!!! no privacy issues!!
- It is light weight with less parameters (Tried lots of LLM and the less parameter the less RAM is required)
Now you can put the raspberry pi 4 to your car and build your own knight rider!!
First step success!! ^^/
Here is the key-points memo
- Use Raspberry Pi Imager and it save tons of time ! Choose option to setup ssh automatically
- Ollama is used
github.com/ollama/ollama
- They key is not choose a heavy parameters model, but with balancing with what you need.
With Trial and error Lizma3.2:1b works best for me
- Custom instruction to make everything shorter to save the process time
github.com/ollama/ollama/blob/main/docs/modelfile.md
Cheers
I trained my own AI from scratch to try to land a rocket. I modified the neural network to achieve a decent result.
This is definitely a really HARD problem that the SpaceX team has solved!!! 🥳
One step forward for human history!
Congrats SpaceX!
Rezipped one-click installer in free patreon post:
patreon.com/posts/5mins-animal-and-135300426
Git Repo:
github.com/KwaiVGI/LivePortrait
0:00 Demo
0:13 Tutorial Start with QR Code
2:19 Start Program
2:44 Choose image and driving video
4:38 Result
I used Google Meet to demonstrate it could stream real-time and should be able to use for video meeting and streaming.
I used Live Portrait and I did a basic tutorial video at here
youtu.be/HU2M7K57ZK0
Now this method also support video2video natually.
Note: I used a different method for making it real-time. It is really hardware heavy, so I am not decided to make any tutorial video for this yet.
"Metahuman Facial Expressions and Lipsync with LivePortrait Tutorial"
youtu.be/HU2M7K57ZK0
Trying to demostrating the potential of subsitute the actual person to Avatar like Metahuman, Anime style Avatar or others.
This tech can apply to Metahuman pretty nice. It could also apply for Virtual Avatara or Animals but not gurantee all would works.
Metahuman Creator
metahuman.unrealengine.com
Live Portrait
github.com/KwaiVGI/LivePortrait
Git
git-scm.com
0:00 Introduction
1:29 Part 1 Prepare and Setup the environements
1:39 Install Miniconda
3:42 Go to the LivePortrait Git
6:18 Download Pretrained Weights
7:44 Part 2 Use AI to add Facial Expression and Lipsync
10:00 Metahuman Creator
13:44 Result 1 Metahuman facial movements
14:45 Result 2 Metahuman Sing with shoulder movements
15:36 Other Results
これは配信や収録で使えそうですね! #lipsync #anime #avatar
Combining the power of #GaussianSplatting and #nerf, this could be really useful for virtual production!
This can also potentially use as backgroudn for MetaHuman!
Made With #UnrealEngine5 #LumaAI
In this video, you would need: (with Links)
- CC4 installed with headshot 2 Pro
reallusion.com/character-creator/headshot
30-day Trial , right now is Black Friday 50% off
- Unreal Engine 5.3 installed
- Download Metahuman Lighting
unrealengine.com/marketplace/en-US/product/mhc-lighting-presets
- Download Auto Setup for Unreal Engine
reallusion.com/auto-setup/unreal-engine/download.html
- Photoshop or GIMP or etc.
- Ref pics Front and Side
- Metahuman Creator
metahuman.unrealengine.com
0:00 Introduction
1:35 Outline and Preparations
5:23 Character Creator 4
6:30 Using Headshot 2
7:58 Image matching tool
8:39 Scruipt Morph
11:17 Side Morph adjustment Method 1
11:55 Side Morph adjustment Method 2 (Recommend)
13:07 Method2: Scale with Photoshop
15:37 Hint on Adjusting
20:14 Done Mesh
20:34 Test with Face motion
21:02 Fix the hair on the head
22:03 Mask Setting
23:33 Edit Face Diffuse in Photoshop
25:53 Fixed Face Texture
26:05 Change Skin to be more Realistic and save in 4k(Optional)
28:59 Make up
29:57 Done Makeup!
30:03 Mesh Adjustment
30:58 Export Head
32:13 Create Unreal Project
32:48 Turn on Metahuman plugin and restart
33:40 Install Auto Setup to Project
35:04 Import Model from CC4
35:58 Successfully Imported
37:43 Create Metahuman from CC4 Mesh
39:48 Check Mesh before upload
40:16 Mesh to Metahuman
40:48 Open Metahuman Creator in Browser
42:40 Import created Metahuman to Project
43:43 Apply CC Texture
45:39 DONE!!!
If it helps a bit pls like and subscribe :)
This is a test to see can I generate my own voice and potentially apply this tech to different scenerio.
The song is eternal wind from Gundam F91
youtu.be/8RbWInyilUU
#metahuman #virtualhuman #ue5 #メタヒューマン #アニメーター
Thanks to Unreal crazy innovation, now we can create animation way faster and easier with the new Metahuman Animator!!!
This video is the tutorial for the whole process.
Also previous video of creating your own mesh to metahuman
youtu.be/htBEU-cJSNg
Feel free to skip to any chapters if needed
0:00 Intro
1:05 Prepare your project
3:34 Capture with you iPhone Live Link Face App
5:42 Setup to reference your downloaded recorded takes
6:06 Process to import with Capture Manager
6:50 Metahuman Identity
8:41 Critical Promote frames process
10:23 View and Compare
10:52 Mesh to Metahuman
11:44 Add teeth
12:18 Turn off lower resolves
12:53 Fit teeth
13:11 Prepare for Performance
13:45 increase Cache
14:29 Check on the new Mesh created by footage
14:46 Metahuman Performance
15:21 Find the region for your performance
16:56 Start Processing
18:20 Check Processed Result
18:52 Export Level Sequence
19:33 Export Animation Sequence
20:22 Create or use the Minimal Default Map
20:31 Use your Metahuman (or use default metahuman)
21:02 Setup sequence
22:36 Add Actor to sequence
24:59 Final Result
Ref Docs:
unrealengine.com/en-US/blog/delivering-high-quality-facial-animation-in-minutes-metahuman-animator-is-now-available
dev.epicgames.com/documentation/en-us/metahuman/metahuman-facial-performance-capture-guidelines
#metahuman #ue5 #animator
Video used:
youtu.be/w4cWE4CYd84.
Really excited for episode 7!!
This video is testing on the new workflow of create content
For fun, learning and education purposes
#推しの子 #推しの子ダンス #びえヨン #Oshinoko #pieyon #wonderdynamics
This is a test for the new workflow of making anime involving ai in the process.
For learning and education purposes
#ai #stablediffusion #sd #anime #推しの子
Workflow details: reddit.com/r/StableDiffusion/comments/13angav/arima_kana_oshinoko_dance_video/?utm_source=share&utm_medium=web2x&context=3
original dance video credit: kanzakianna36
↓HDRP project link, recommend to use 2023.1.0b9 Unity version (at the time I recorded this at 5/6/2023)
assetstore.unity.com/packages/essentials/tutorial-projects/enemies-hdrp-248975
Unity version download link:
unity.com/releases/editor/beta/2023.1.0b9
Some of the highlighted points:
4k rendering with DLSS
Real-time ray traced reflicection, AO.
Realistic skin system
Better simulation on Hair
The other Hair Demo project
github.com/Unity-Technologies/com.unity.demoteam.hair
Unity GDC 2022 Demo of Digital Human
youtube.com/watch?v=tz3dpl7FlH4&ab_channel=Unity
More details:
unity.com/demos/enemies
I think this test looks pretty good :)
#wonderdynamics #dev #anime #vtuber
This is a comparision for a new workflow of making anime.
- Original
- Protype
- Before Denoise
- Final
I basically able to turnt myself to a completely different person and to a Demon Slayer character.
Full video
youtu.be/6O2Kc7Y2AW4
For learning and education purpose
Tutorial later if people showed interest :)
#ai #demonslayer #anime #artificialintelligence #stablediffusion #tutorial #crunchyroll #kimetsunoyaiba
This is a comparation of the test for a new workflow of making anime.
I basically able turnt myself into anime Demon Slayer character and able to minized the noise on the character
Full Video
youtu.be/6O2Kc7Y2AW4
For learning and education purpose
#ai #demonslayer #anime #artificialintelligence #stablediffusion #shorts #tutorial
This is a test for a new workflow of making anime.
I basically turnt myself into Demon Slayer character.
For learning and education purpose
Tutorial later
#ai #demonslayer #anime #artificialintelligence #stablediffusion #demonslayerusa #crunchyroll
You can see how the gundam dynamically interacting with the lighting in the background.
#NeRFs viewer in Unreal is insane!!
- Combining NeRFs content in virtual production and traditional game
- It support PC VR!!!
- It could interact with Nanite meshes
This definitely is the future! Now we finally can used what produced from NeRFs to our games!
More details ↓
docs.lumalabs.ai/9DdnisfQaLN1sn
Last but not least!
俺達が!ガンダムだ !
#AI #ArtificialIntelligence #LumaAI #UE5 #UnrealEngine5 #Gundam
Japan link
amazon.co.jp/dp/B081K3MPFQ
Prerequisite Links
Mocopi Setup (Skip the VRChat part)
youtu.be/PfOOLX6Gi6Y
Mesh to Metahuman
youtu.be/htBEU-cJSNg
Or Metahuman Creator (Old)
youtu.be/-cfk5dsC2Rc
This video is usingUnreal Engine 5.1
Bonemaps (credits to Rokoko Studio)
drive.google.com/file/d/1Xjo3_gPqOImMCC3K99Y9gcR11gCH3JwF/view?usp=sharing
rokoko.com/products/studio
Chapters
0:00 Introduction
1:05 Prepare the project
1:21 Install Rokoko Studio LIVE
1:38 Create Empty Project
2:43 Turn on the Rokoko Plugin
3:56 Import your Metahuman
5:53 Put BoneMap Folder to Content Folder(See Description link)
8:34 Back to Project at Unreal 5.1
9:20 Metahuman Blueprint
9:53 LODSync
10:08 Finding Metahuman animgraph
12:19 Editing the animgraph
15:35 Fixing mouth problem
18:11 Main Blueprint editing
21:14 It's Morphing time!
21:51 Connect Rokoko Studio
27:13 Setup Rokoko in Map
27:59 Confusing Activate button
29:08 Live LInk Window
29:34 Add Metahuman to Map
29:53 Set animebp and fix for face
30:44 Calibration Again
31:17 DONE!
#mocopi #sony #metahuman #vtuber #unrealengine
Mocopi intro and tutorial, full body in VRChat
youtu.be/PfOOLX6Gi6Y
Make your own Metahuman tutorial
youtu.be/htBEU-cJSNg
#mocopi #shorts #sony #unrealengine #metahumans
Today 1/20/2023, Sony Japan just released a brand new motion capture sensors called "Mocopi"
sony.jp/mocopi
It is really small and support output to video, bvh, vrm and real time streaming.
The first part of the tutorial would explain how to set the sensors up, wear it, use the app.
The second part of the tutorial would explain how to set up in VRChat and the potential problem you might encounter.
0:00 Introduction
1:02 How to setup your mocopi sensors
4:45 Calibration (Could be Tricky)
6:01 Introducing Mocap App (which is released in japan app store)
8:09 VRChat
8:41 Enable OSC in VRChat
9:05 Allow VRChat in Firewall Setting
9:20 Steps on sending OSC VRChat Signal from the mocopi app
10:39 DONE
10:45 Fixing problem of Body and legs are not detected
11:05 Recommend to use mocopi default avatar for VRChat full body tracking usage
11:28 Test on short Avatar (KiraDragon)
11:55 Test on Avatar from MakeAvatar App
Iphone Japan app store link: apps.apple.com/jp/app/mocopi/id6444393701
Note
Q VRChat Calibrate button did not shown
A There could be couple reasons, make sure you checked again at 9:20 and pause to compare the screen, make sure it top right second button is send instead of save. Good luck!
Thanks for watching :)
#mocopi #vrchat
sony.jp/mocopi
You can see the demo video and understand how the calibration works and see how convienent this is!
Under my test, it works better on my iPhone 12 Pro Max then on my Galaxy Fold 3.
Galaxy Fold 3 seems to have around 2 sec delay.
The app comes with Mocopi support video, bvh format and supports #vrchat !
Detail full tutorial
youtu.be/PfOOLX6Gi6Y
Cheers
If there is enough people interested in, I would probably do a tutorial on this :)
For learning and educational purpose
This is a Tutorial to Scan your own custom face, aiming to create a realistic face of yourself to the Unreal Engine #Metahuman format.
Hopefully you would be able to use your own metahuman with you own face after this tutorial. The metahuman plugin still have space to improve, but it is really really IMPRESSIVE!
Thank you #Trnio Plus for such great quality, and #UnrealEngine for this magical Tech.
0:00 Intro
0:59 Trnio Plus(IPhone Only)
3:01 Few things to be aware for your pictures
6:21 Edit Orientation in Trnio Plus
9:50 Where to Upload zipped Obj file
10:23 Setup Metahuman Plugin
12:56 Import the Obj file
14:13 Using MetaHuman Plugin
16:04 AutoTrack Face
16:38 MetaHuman Identity Solve
18:12 Select Body and Upload
19:20 MetaHuman Creator
23:00 DONE and Compare!
Trnio Plus App
apps.apple.com/us/app/trnio-plus-3d-scanner/id1546160586
Metahuman Plugin
unrealengine.com/marketplace/en-US/product/metahuman-plugin
Metahuman Creator
metahuman.unrealengine.com
Ref
docs.metahuman.unrealengine.com/en-US/mesh-to-metahuman-in-unreal-engine
youtu.be/xepZ_HJ7X5Y
Ref Docs
unrealengine.com/en-US/blog/new-release-brings-mesh-to-metahuman-to-unreal-engine-and-much-more
docs.metahuman.unrealengine.com/en-US/mesh-to-metahuman-quick-start
Anime Ya Boy Kongming! OP ChiKiChiKiBanBan ft. KiraDragon & KikyoPanda
The motion data and op video is released free with avex picture over here
youtu.be/2ZhPmqx1BSY
and motion data is here
https://paripikoumei.booth.pm/items/3872754
You can create by yourself too!!
It's Party time!!
#パリピ孔明
#チキチキバンバン
#踊ってみた
#YaBoyKongming
#Chikichikibanban
#dancecover
#3dcg
Current Version Unreal Engine 5.0.1
0:00 Intro
0:59 Explanations
6:40 Getting the project ready
8:32 Import Metahuman
11:02 Import ThirdPerson
11:41 Retargeting between Metahuman and Mannequin
12:47 Using Compatible Skeleton (See UE5 Release Notes)
14:19Comparing Metahuman and Mannequins Animation
15:33 Metahuman First Test
16:28 Testing between Direct and Retarget Animation
19:13 Mixamo Retargeting and Disclaimer
20:39 Import Mannequin to Mixamo and download
22:17 Convert Mixamo animation to Unreal Engine
23:17 Import Animation to Unreal Engine
25: 25 Retarget Animation from UE4 Mannequin to UE5
26:48 Testing Mixamo Animation on Metahuman
DONE!!!
Ref Links
Ik-rig-animation-retargeting
docs.unrealengine.com/5.0/en-US/ik-rig-animation-retargeting-in-unreal-engine
Skeleton compatibility
docs.unrealengine.com/5.0/en-US/unreal-engine-5-0-release-notes
Mixamo Converter
https://terribilisstudio.fr/
New UE5 MetaHuman Fly as Superman Tutorial
youtu.be/8Vxot0JU_kU
Old video that is renewed by this video
UE5 Early Acess (EA):
Simple Metahuman with Mixamo Animation Tutorial with Third Person Controller Pt.2
youtu.be/JsTw3sAor-U
This Gym is in Japan Tokyo and you can reserve over here!
machbody.com
Golden Week 2022! 一日バク転チャレンジ!!
予約↑のリンクで^^
The other cleaner way of assign skeleton by using a new feature called compatible skeleton
youtu.be/tGU517bpX2g. On12:47
if you had torso, leg, feet, etc misaligned, you might need to reset your location or rotation after you copy and paste to your new blueprint.
If your rendering seems to be odd in the editor(like leg moving with curve) , try to use scalability high setting
If you try to build your metahuman in this project, you will need to turn on the "Apple ARKit" and "Apple ARKit Face Support" in the plugins. Or lese your project will crashed.
If you try to build your metahuman hair in this project, to make you hair visible, you will need to turn on the "Alembic Groom Importer" in in the plugins.
Built exe for testing, 3rd person movement works
dropbox.com/t/Xxq244yV50KdhlPF
↑Note the above file metahuman hair simulation is not on
Unreal engine version in this tutorial is UE5.0.1
This is a tutorial intended to update the old video on the same topic of applying Metahuman to the third person.
0:00 Highlight
0:07 Intro
0:45 Preparing the Project
2:27 Import MetaHuman
3:49 Fixing Metahuman LOD
5:33 Create Metahuman Third Person Character
7:50 Assign Skeleton to Metahuman
9:16 Fix Torso and shoes not following the Body (Master Component)
11:43 First Test
12:01 Fix Ice Skating Problem (Control Rig)
12:27 DONE
Ref
Previous UE4 and UE5 Preview Simple Metahuman and Third Person Tutorial
youtu.be/TrIvO9fA7RY
Previous Video on Metahuman Creator created my Monkey King Character
youtu.be/-cfk5dsC2Rc
Feel free to skip to any part of the video through the timeline feature.
This tutorial also included a quick and dirty way retargeting from the animation assets to the Metahuman.
The other cleaner way of assign skeleton by using a new feature called compatible skeleton
youtu.be/tGU517bpX2g. On12:47
Before you start the tutorial, it is highly recommended that you spend time to install the new UE5.0, City Sample and Flight Animation(NOT FREE) used in this tutorial.
0:00 Demo of final highlight result of this tutorial
1:00 Intro
2:06 Download City Sample (see links at bottom)
2:50 Get Flight Animation Asset and create the dummy project(see links at bottom)
4:56 Open City Sample Project
5:59 Load City to check the scalability that fits your PC
6:42 Import Metahuman
8:56 Fix LODs on hairs
9:29 Switch game mode
12:01 First Test
12:34 Overide World Settings
13:14 Second Test
13:40 Get Inputs
16:38 Third Test
18:24 Metahuman Setup
22:23 Retargeting (Quick and Dirty Way)
26:26 Final Result (DONE!)
City Sample
unrealengine.com/marketplace/en-US/learn/city-sample?lang=en-US
Flight Animation that used in this tutorial
marketplace-website-node-launcher-prod.ol.epicgames.com/ue/marketplace/en-US/product/superhero-flight-animations
If you like it or if it helps you a bit, please press like and subscribe :)
My PC is really weak, so I changed the settings in the scalability to medium, hence able to move smoothly for testing. It is still quite lag in the editor.
There are a lot of changes in the latest UE5, tutorial of this will be come out soon.
City Sample Project
unrealengine.com/marketplace/en-US/product/city-sample
Previous similar project
Valley Of The Ancient
youtu.be/7J6lEdeAnE0
Source and author from JimWest
Please feel free to skip to any section you see fit by clicking on the time
0:00 Intro
2:55 Release page from github
3:58 Command line
5:24 Testing on Webcam input
6:50 Found Webcam
8:18 LiveLink settings
9:20 DONE
It will works on metahuman sample project
youtu.be/erwXtT9lXR4
Links
Download link
github.com/JimWest/MeFaMo/releases
Project github
github.com/JimWest/MeFaMo
Google Media Pipe
github.com/google/mediapipe
Alternative Better quality but not free, Faceware
youtu.be/fwAjwA44mSk
Want to have the exact same character from the video
UE5 MetaHuman Change Clothes
youtu.be/x37uIyt7K2Y
Create your own MetaHuman
youtu.be/-cfk5dsC2Rc
Result is not Perfect, but with enough data, it should works even better.
Method 1 Video
youtu.be/xepZ_HJ7X5Y
Method 2 Original Video from the Metahuman Beginner Series
youtu.be/-cfk5dsC2Rc
Metahuman Beginner Series Playlist
youtube.com/playlist?list=PLXdLPBqF9snwxPJ2WYu_1awHxqUNkJ2Iy
New Series of Tutorial will be out later (After preparation is done)
0:00 Acquire the human face
0:08 Testing on Blendshape and Topology
0:35 Testing on Ziva ZRT
New Series of Tutorials will be out later
Tokyo Game Show VR 2021 Main sites
tgsvr.com/en
Download page
tgsvr.com/en/download
Door (Browser based VR that works on PC and Mobile)
https://door.ntt/
Door Event Entrance
https://door.ntt/web/tgsvr2021_sky/index.html
Events dates
from 9:00 a.m.(JST) on September 30
to 2:00 a.m.(JST) on October 4, 2021
0:00 Intro for this session of the channel
1:03 Tokyo Game Show Explain
2:07 Two kind of platforms Explain
2:30 #DOORNTT Explain
3:24 Ending soon (JOIN QUICK!)
4:16 Social Experience in VR (#xambr and DOOR)
4:50 Highlight Video start with Quest 2 Experience
5:23 Game Float Sky Entrance (DOOR)
6:40 Browser VR Social Experience
7:50 Game Float (xambr)
8:30 Group pic in main Entrance
10:19 Monster Hunter Booth
11:03 Attack on Titans Booth
12:55 Macross Frontier (マクロスF ) Booth
13:09 Sneak peak on Concert
Hope you guys enjoy :)
#TGS #TGSVR #TGSVR2021 #TokyoGameShow
This is acting as a short summary so far for the milestones.
Links to learn to get into this point
Simple Tutorial on using the sample metahuman with face and head recognitions
youtu.be/erwXtT9lXR4
Create your own with MetaHuman #Creator
youtu.be/-cfk5dsC2Rc
Use with Third Person Controller Pt. 1
youtu.be/TrIvO9fA7RY
Use with Third Person Controller with #Mixamo Pt. 2
youtu.be/JsTw3sAor-U
Change your character clothing
youtu.be/x37uIyt7K2Y
Demo project #ValleyOfTheAncient
youtu.be/7J6lEdeAnE0
youtu.be/tGU517bpX2g
This Unreal Engine Early Access(EA) Tutorial compose with 4 parts,
1. Review on how to initial the third person metahuman controller in #UnrealEngine5 (#UE5)
2. Note on mapping the parts together for #Metahuman
3. #Mixamo #Animation Retargetting
4. Put the Mixamo Animation to the third person controller
Previously
Simple Metahuman and Third Person Tutorial Pt.1
youtube.com/watch?v=TrIvO9fA7RY&t=5s&ab_channel=RelaxKiral
0:00 Intro
0:13 Note on Previous Third Person Controller Video
1:44 Quick Review on Setup Metahuman Third person project in Unreal
3:33 Note and showing the actual construction graph for Metahuman
4:34 MIXAMO HERE!!! (if you already know all the basic, skip to here! )
6:32 Convert Mixamo to unreal animation
7:33 Import your converted Mixamo Animation and show the problem
10:35 Import Mixamo Animation with Offset this time
12:54 Swap Run Animation to Mixamo Animation (#Naruto / Ninja Run)
14:04 DONE!
Reference docs
Fixing Modular Parts of Metahuman
docs.unrealengine.com/4.26/en-US/AnimatingObjects/SkeletalMeshAnimation/WorkingwithModularCharacters
Mixamo
mixamo.com
Mixamo Converter
http://terribilisstudio.fr/?section=MC
Details on RootMotion setting
docs.unrealengine.com/4.26/en-US/AnimatingObjects/SkeletalMeshAnimation/RootMotion
0:00 Demo and Intro
1:04 Ready for the Project
3:10 1. Import MetaHuman (LODSync 0) (Previous video on creating MetaHuman youtu.be/-cfk5dsC2Rc )
4:40 2. Enable ARKit and Face Capture Plugin
6:14 3. Import The Target Character(My case is Wukong)
7:54 Get the target character ready
8:29 1. Disable the target character head related parts
11:37 2. Fix cannot export character texture problem (if needed)
25:11 Retarget Skeleton
26:44 Get the metahuman ready
27:24 1. Remove all metahuman clothing
27:45 Combined MetaHuman and target character
27:52 1. Assign Metahuman body to target character body
28:37 2. Fix Body Material
29:02 3. Fix Pose Problem
29:55 4. Fix Face Material (Note you can adjust position of Face by moving its position)
36:13 Quick Result
36:55 Test with LIVE Link (Previous Tutorial on LIVE Link youtu.be/erwXtT9lXR4 )
37:17 Shadow clone
Reference link:
docs.unrealengine.com/4.26/en-US/RenderingAndGraphics/Materials/Functions/Reference/Gradient
MetaHuman Beginner Series (EN) List:
youtube.com/playlist?list=PLXdLPBqF9snwxPJ2WYu_1awHxqUNkJ2Iy


