DeepAI AI Chat
Log In Sign Up

How well can Text-to-Image Generative Models understand Ethical Natural Language Interventions?

10/27/2022
by   Hritik Bansal, et al.
0

Text-to-image generative models have achieved unprecedented success in generating high-quality images based on natural language descriptions. However, it is shown that these models tend to favor specific social groups when prompted with neutral text descriptions (e.g., 'a photo of a lawyer'). Following Zhao et al. (2021), we study the effect on the diversity of the generated images when adding ethical intervention that supports equitable judgment (e.g., 'if all individuals can be a lawyer irrespective of their gender') in the input prompts. To this end, we introduce an Ethical NaTural Language Interventions in Text-to-Image GENeration (ENTIGEN) benchmark dataset to evaluate the change in image generations conditional on ethical interventions across three social axes – gender, skin color, and culture. Through ENTIGEN framework, we find that the generations from minDALL.E, DALL.E-mini and Stable Diffusion cover diverse social groups while preserving the image quality. Preliminary studies indicate that a large change in the model predictions is triggered by certain phrases such as 'irrespective of gender' in the context of gender bias in the ethical interventions. We release code and annotated data at https://github.com/Hritikbansal/entigen_emnlp.

READ FULL TEXT

page 10

page 11

page 12

06/02/2021

Ethical-Advice Taker: Do Language Models Understand Natural Language Interventions?

Is it possible to use natural language to intervene in a model's behavio...
02/08/2022

DALL-Eval: Probing the Reasoning Skills and Social Biases of Text-to-Image Generative Transformers

Generating images from textual descriptions has gained a lot of attentio...
11/17/2022

Is the Elephant Flying? Resolving Ambiguities in Text-to-Image Generative Models

Natural language often contains ambiguities that can lead to misinterpre...
11/09/2015

Generating Images from Captions with Attention

Motivated by the recent progress in generative models, we introduce a mo...
01/18/2021

Teach me how to Label: Labeling Functions from Natural Language with Text-to-text Transformers

Annotated data has become the most important bottleneck in training accu...
11/17/2022

GLAMI-1M: A Multilingual Image-Text Fashion Dataset

We introduce GLAMI-1M: the largest multilingual image-text classificatio...
02/12/2021

They, Them, Theirs: Rewriting with Gender-Neutral English

Responsible development of technology involves applications being inclus...