BiPC: Bidirectional Probability Calibration for Unsupervised Domain Adaption

TL;DR

BiPC introduces a bidirectional probability calibration method that enhances unsupervised domain adaptation by leveraging pre-trained classifiers and task-specific heads, applicable to diverse neural network architectures.

Contribution

The paper proposes a novel bidirectional probability calibration framework for UDA, integrating probability alignment and Gini impurity loss to improve adaptation across different network types.

Findings

BiPC achieves state-of-the-art results on multiple UDA benchmarks.

The method effectively calibrates probabilities, boosting model robustness against domain shifts.

Applicable to CNNs and Transformers, demonstrating broad versatility.

Abstract

Unsupervised Domain Adaptation (UDA) leverages a labeled source domain to solve tasks in an unlabeled target domain. While Transformer-based methods have shown promise in UDA, their application is limited to plain Transformers, excluding Convolutional Neural Networks (CNNs) and hierarchical Transformers. To address this issues, we propose Bidirectional Probability Calibration (BiPC) from a probability space perspective. We demonstrate that the probability outputs from a pre-trained head, after extensive pre-training, are robust against domain gaps and can adjust the probability distribution of the task head. Moreover, the task head can enhance the pre-trained head during adaptation training, improving model performance through bidirectional complementation. Technically, we introduce Calibrated Probability Alignment (CPA) to adjust the pre-trained head's probabilities, such as those from an ImageNet-1k pre-trained classifier. Additionally, we design a Calibrated Gini Impurity (CGI) loss to refine the task head, with calibrated coefficients learned from the pre-trained classifier. BiPC is a simple yet effective method applicable to various networks, including CNNs and Transformers. Experimental results demonstrate its remarkable performance across multiple UDA tasks. Our code will be available at: https://github.com/Wenlve-Zhou/BiPC.