Localization and manipulation of immoral visual cues for safe text-to-image generation
Current text-to-image generation methods produce high-resolution and high-quality images,
but they should not produce immoral images that may contain inappropriate content from the …
but they should not produce immoral images that may contain inappropriate content from the …
Zero-shot visual commonsense immorality prediction
Artificial intelligence is currently powering diverse real-world applications. These
applications have shown promising performance, but raise complicated ethical issues, ie …
applications have shown promising performance, but raise complicated ethical issues, ie …
Improving identity-robustness for face models
Q Qi, S Ardeshir - arXiv preprint arXiv:2304.03838, 2023 - arxiv.org
Despite the success of deep-learning models in many tasks, there have been concerns
about such models learning shortcuts, and their lack of robustness to irrelevant confounders …
about such models learning shortcuts, and their lack of robustness to irrelevant confounders …
On bias and fairness in deep learning-based facial analysis
Facial analysis systems are used in a variety of scenarios such as law enforcement, military,
and daily life, which impact important aspects of our lives. With the onset of the deep …
and daily life, which impact important aspects of our lives. With the onset of the deep …
LLM2Loss: Leveraging Language Models for Explainable Model Diagnostics
S Ardeshir - arXiv preprint arXiv:2305.03212, 2023 - arxiv.org
Trained on a vast amount of data, Large Language models (LLMs) have achieved
unprecedented success and generalization in modeling fairly complex textual inputs in the …
unprecedented success and generalization in modeling fairly complex textual inputs in the …