AI / ML · 2024
Emotion Translation with Transformers
A transformer model that rewrites the emotion of a sentence while preserving its underlying meaning.
Overview
We developed a model that can change the emotion of a particular statement without changing the context of the sentence — taking a neutral piece of text and re-rendering it as joyful, angry, or melancholic while keeping its factual meaning intact. The project explores controllable text generation and the boundary between sentiment and semantics.