Learnable Adaptive Cosine Estimator (LACE) for Image Classification
Abstract
In this work, we propose a new loss to improve feature discriminability and classification performance. Motivated by the adaptive cosine/coherence estimator (ACE), our proposed method incorporates angular information that is inherently learned by artificial neural networks. Our learnable ACE (LACE) transforms the data into a new "whitened" space that improves the inter-class separability and intra-class compactness. We compare our LACE to alternative state-of-the art softmax-based and feature regularization approaches. Our results show that the proposed method can serve as a viable alternative to cross entropy and angular softmax approaches. Our code is publicly available.
Cite
Text
Peeples et al. "Learnable Adaptive Cosine Estimator (LACE) for Image Classification." Winter Conference on Applications of Computer Vision, 2022.Markdown
[Peeples et al. "Learnable Adaptive Cosine Estimator (LACE) for Image Classification." Winter Conference on Applications of Computer Vision, 2022.](https://mlanthology.org/wacv/2022/peeples2022wacv-learnable/)BibTeX
@inproceedings{peeples2022wacv-learnable,
title = {{Learnable Adaptive Cosine Estimator (LACE) for Image Classification}},
author = {Peeples, Joshua and McCurley, Connor H. and Walker, Sarah and Stewart, Dylan and Zare, Alina},
booktitle = {Winter Conference on Applications of Computer Vision},
year = {2022},
pages = {3479-3489},
url = {https://mlanthology.org/wacv/2022/peeples2022wacv-learnable/}
}