Cross-Domain Collaborative Normalization via Structural Knowledge

Authors

  • Haifeng Xia Tulane University
  • Zhengming Ding Tulane University

DOI:

https://doi.org/10.1609/aaai.v36i3.20181

Keywords:

Computer Vision (CV)

Abstract

Batch Normalization (BN) as an important component assists Deep Neural Networks in achieving promising performance for extensive learning tasks by scaling distribution of feature representations within mini-batches. However, the application of BN suffers from performance degradation under the scenario of Unsupervised Domain Adaptation (UDA), since the estimated statistics fail to concurrently describe two different domains. In this paper, we develop a novel normalization technique, named Collaborative Normalization (CoN), for eliminating domain discrepancy and accelerating the model training of neural networks for UDA. Unlike typical strategies only exploiting domain-specific statistics during normalization, our CoN excavates cross-domain knowledge and simultaneously scales features from various domains by mimicking the merits of collaborative representation. Our CoN can be easily plugged into popular neural network backbones for cross-domain learning. On the one hand, theoretical analysis guarantees that models with CoN promote discriminability of feature representations and accelerate convergence rate; on the other hand, empirical study verifies that replacing BN with CoN in popular network backbones effectively improves classification accuracy in most learning tasks across three cross-domain visual benchmarks.

Downloads

Published

2022-06-28

How to Cite

Xia, H., & Ding, Z. (2022). Cross-Domain Collaborative Normalization via Structural Knowledge. Proceedings of the AAAI Conference on Artificial Intelligence, 36(3), 2777-2785. https://doi.org/10.1609/aaai.v36i3.20181

Issue

Section

AAAI Technical Track on Computer Vision III