Name: |
|
Train NN classifying emotional charge of pictures |
You will learn: |
|
machine learning, image analysis, emotions |
Difficulty: |
|
moderate |
Excitement: |
|
Learn how to process images, understand emotions. |
Description: |
|
Neural networks are state of the art approach to classifying pictures. They can recognize whether there is a cat on the picture or follow a line on the road steering google cars. On the other hand cognitive sciences are a fast growing area. Research in emotions often requires large corpuses of emotionally annotated data. Could we train a neural network that could automatically assess the emotional charge (valence and arousal) of input photograph?
|
Further Reading: |
|
|
How to sign up: |
|
Information [[Bachelor and Master thesis proposals| here]] .
|