Scribd
Upload
11 views

Assignment k4

Uploaded by

wwcfnm5mcy
Copyright
© © All Rights Reserved
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
11 views

Assignment k4

Uploaded by

wwcfnm5mcy
Copyright
© © All Rights Reserved
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
You are on page 1/ 2

ASSIGNMENT

Q1)Yes, AI can play games including the HTML5 games like the ones on website
that is provided by you. It can play a wide range of games.
>Different techniques that can be used are:
-Game state analysis: for analyzing current state of game.

-Decision Making: Uses reinforcement(like Deep Q networks) and supervised


learning(by training AI using labeled data of game states) techniques.

-Game interaction: Uses tools such as selenium or PyAutoGUI or also uses web
Scarping and Browser Automation.

For computer Vision to play games that focuses on letting computers to clarify
and acknowledge visual information from videos and images. For it you would
need:
Capture game screen, object detection ,image processing etc.

>Tools that need to be used for incorporating computer vision to games are:
-Python(for script)
-OpenCV(for CV tasks)
-TensorFlow(for building and training neural networks)
-Pillow(for image processing)
and there are many others available as well.

Q2)Yes, AI animation is possible and has become rapidly growing field with
advancements that are much sophisticated in AI and Machine Learning.
The video provided for reference shows AI-Generates animations.

> Certain AI/ML tools that can be used to develop animation are:
-Deep learning animation tools like Adobe Fresco, Prisma to use deep learning
algorithms
-Computer vision Animation tools like Adobe After Effects, OpenCV to track and
animate objects in videos.

-DeepArt is a tool that applies artistic styles to images and videos using neural
networks.

-Neural Frames offers AI powered animation generator for crafting animations


and music videos.
With growing popularity new tools are being developed everyday.

> To develop some basic tools for the same we can:


1.Gather Data- gathering datasets of images and videos that will be used to train
the model.

2.Preprocessing the data -to clean and get more refined data that can be useful
to us we need to preprocess it.

3.Choose a Model-Selecting an pertinent model architecture like CNNs, GANs,


RNNs.

4.Train and Test the model-After refining and choosing a certain model type one
needs to train the model by using different AI/ML algorithms that the person
wants to add using Pytorch etc, after the training bit the model is tested to check
if any further improvements are to be made

5.Deployment of model: Once the model is trained and tested deploy it using a
fitting platform. Combine The model with other tools to enhance it even further
for user ease.

You might also like