آموزش سیستم AI Vision هوش مصنوعی در Houdini

عنوان دوره : cmiVFX – Houdini AI Vision System

سطح : متوسط

زمان کل دوره : 3 ساعت

تعداد فایل های تصویری : 42

سال آموزش : 2020

فایل تمرینی : دارد

مربی دوره : Sadjad Rabiee

نرم افزار : Houdini

زیرنویس انگلیسی : ندارد

cmiVFX – Houdini AI Vision System

cmiVFX has released its first Artificial Intelligence Brain solution for SideFX Software’s Houdini Animation Software Package. This video is a milestone in cmiVFX history, for it is currently the only one of its kind. Our mentor delves deep into the virtual psyche of the Visual Effects Artist’s mind by preparing a digital brain solution for a car animation with obstacles. Naturally, since this process is procedural, the end users can take apart this system modularly and build their own characters movements and skills to follow the needs of their workflow. This video will shock some of you long time Houdini users with some methods not commonly seen in the industry. When it comes to Houdini training, you know cmiVFX has something new for you when and where you want it.. just log in and watch. Short Description In this video you will learn how to set up Houdini to operate like a true Artificial Intelligence Solution by diving into some very fun and overlooked tools often forgotten about by the industry. Houdini can be your answer to much more than just Explosions and Dynamic Simulations? Come learn how to make a virtual brain station!

Vision Node
In the first chapter we create a custom VOP SOP that is called Vision Node. While doing this, we get a perspective view of the agent in real time with this node so that in other later chapters we can take view data from this node to process AI behaviors.

Attributes
In this portion of the video we create some custom attributes that are used for AI system like Behavior attributes and color Attributes to help make our events interesting or stimulating to the eye.

Agent Mover
This chapter is where we create another network that is called the “Agent Mover” that allows the Agents to move and turn with behavior attributes and avoid some simple obstacles.

Car Model
Now that we have a pretty stable system, we import and setup an interesting cartoon car model as well transfer animation data from the Agent Mover network to move and turn this car through out its course.

Improve Animation
This chapter’s intention is to create an interesting, useful Digital Asset that is called JK_Smooth_Value to make the car agent animate more smoothly.

Finalizing
In this step we fix some bugs and take the car’s color from Agent_eye network to apply to car agent character.

Rendering
and finally we create some materials and light the scene and import the cache file from the car animation to render the scene.