Black Sheep in the Herd: Playing with Spuriously Correlated Attributes for Vision-Language Recognition

X Tian, S Zou, Z Yang, M He, J Zhang - arxiv preprint arxiv:2502.15809, 2025 - arxiv.org
Few-shot adaptation for Vision-Language Models (VLMs) presents a dilemma: balancing in-
distribution accuracy with out-of-distribution generalization. Recent research has utilized low …

Bias Mitigation in Visual Models Through Split-Model Regularization: A Case Study in Online Class Participants' Engagement Classification

J Thiering - 2022 - search.proquest.com
Addressing the issue of bias in deep learning models for visual tasks is challenging. Deep
learning models often learn by minimizing training error through observation of spurious …