※ 引述《dharma (达)》之铭言：
: 有人说可能是架设摄影机来观看萤幕画面
: 机械手臂操控键鼠(限制手速)
: 让AI能以最接近人类的方式来竞赛
: 也有人说是游戏内容会转换成给AI软件读取的资讯
: 类似下面影片这样
: https://www.youtube.com/watch?v=5iZlrBqDYPM
: 请问比赛方式有定案了吗
: thanks
不需要可能这个可能那个的乱猜
Deepmind早就解释完了
https://goo.gl/DJgHXv
节录一些重点
An agent that can play StarCraft will need to demonstrate effective use of
memory, an ability to plan over a long time, and the capacity to adapt plans
based on new information. Computers are capable of extremely fast control,
but that doesn’t necessarily demonstrate intelligence, so agents must
interact with the game within limits of human dexterity in terms of “Actions
Per Minute”.
We’ve worked closely with the StarCraft II team to develop an API that
supports something similar to previous bots written with a “scripted”
interface, allowing programmatic control of individual units and access to
the full game state (with some new options as well). Ultimately agents will
play directly from pixels, so to get us there, we’ve developed a new
image-based interface that outputs a simplified low resolution RGB image data
for map & minimap, and the option to break out features into separate “layers
”, like terrain heightfield, unit type, unit health etc. Below is an example
of what the feature layer API will look like.
重点中的重点翻译
1. 这个可以玩星海2的AI要展示的是
a. 有效率的运用内存
b. 长期计画的能力
c. 因应新资讯调整计画的能力
2. AI会被限制在人类可及的APM之内
3. Deepmind跟星海2团队合作
开发可以直接用程式来控制单位以及存取游戏状况的API
4. AI会直接读取像素来玩星海2
为了可以让AI直接读取像素
他们也开发了一套简化的低分辨率RGB的地图以及小地图影像资料
以及可以将影像资料分为其他图层的选项
让AI也能透过其他图层取得地形高低、单位种类、单位血量等等
(那接口大概就会像原PO贴的影片长的样子)