The bias amplification paradox in text-to-image generation
Bias amplification is a phenomenon in which models increase imbalances present in the
training data. In this paper, we study bias amplification in the text-to-image domain using …
training data. In this paper, we study bias amplification in the text-to-image domain using …
Fairness in Autonomous Driving: Towards Understanding Confounding Factors in Object Detection under Challenging Weather
B Pathiraja, C Liu, R Senanayake - arXiv preprint arXiv:2406.00219, 2024 - arxiv.org
The deployment of autonomous vehicles (AVs) is rapidly expanding to numerous cities. At
the heart of AVs, the object detection module assumes a paramount role, directly influencing …
the heart of AVs, the object detection module assumes a paramount role, directly influencing …
Leveraging CLIP for Inferring Sensitive Information and Improving Model Fairness
M Zhang, R Chunara - arXiv preprint arXiv:2403.10624, 2024 - arxiv.org
Performance disparities across sub-populations are known to exist in deep learning-based
vision recognition models, but previous work has largely addressed such fairness concerns …
vision recognition models, but previous work has largely addressed such fairness concerns …
[PDF][PDF] Robust Machine Learning: Detection, Evaluation and Adaptation Under Distribution Shift
S Garg - 2024 - kilthub.cmu.edu
Deep learning, despite its broad applicability, grapples with robustness challenges in real-
world applications, especially when training and test distributions differ. Reasons for the …
world applications, especially when training and test distributions differ. Reasons for the …
Prompting for Robustness: Extracting Robust Classifiers from Foundation Models
Machine learning models can fail when trained on distributions with hidden confounders
(spuriously correlated with the label) and tested on distributions where such correlations are …
(spuriously correlated with the label) and tested on distributions where such correlations are …