Inferring Offensiveness In Images From Natural Language Supervision
share
Probing or fine-tuning (large-scale) pre-trained models results in state-of-the-art performance for many NLP tasks and, more recently, even for computer vision tasks when combined with image data. Unfortunately, these approaches also entail severe risks. In particular, large image datasets automatically scraped from the web may contain derogatory terms as categories and offensive images, and may also underrepresent specific classes. Consequently, there is an urgent need to carefully document datasets and curate their content. Unfortunately, this process is tedious and error-prone. We show that pre-trained transformers themselves provide a methodology for the automated curation of large-scale vision datasets. Based on human-annotated examples and the implicit knowledge of a CLIP based model, we demonstrate that one can select relevant prompts for rating the offensiveness of an image. In addition to e.g. privacy violation and pornographic content previously identified in ImageNet, we demonstrate that our approach identifies further inappropriate and potentially offensive content.
READ FULL TEXT