-
[pdf]
[bibtex]@InProceedings{Majurski_2024_CVPR, author = {Majurski, Michael and Menon, Sumeet and Favardin, Parniyan and Chapman, David}, title = {A Method of Moments Embedding Constraint and its Application to Semi-Supervised Learning}, booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) Workshops}, month = {June}, year = {2024}, pages = {7809-7818} }
A Method of Moments Embedding Constraint and its Application to Semi-Supervised Learning
Abstract
Discriminative deep learning models with a linear+softmax final layer have a problem: the latent space only predicts the conditional probabilities p(Y|X) but not the full joint distribution p(YX) which necessitates a generative approach. The conditional probability cannot detect outliers causing outlier sensitivity in softmax networks. This exacerbates model over-confidence impacting many problems such as hallucinations confounding biases and dependence on large datasets. To address this we introduce a novel embedding constraint based on the Method of Moments (MoM). We investigate the use of polynomial moments ranging from 1st through 4th order hyper-covariance matrices. Furthermore we use this embedding constraint to train an Axis-Aligned Gaussian Mixture Model (AAGMM) final layer which learns not only the conditional but also the joint distribution of the latent space. We apply this method to the domain of semi-supervised image classification by extending FlexMatch with our technique. We find our MoM constraint with the AAGMM layer is able to match the reported FlexMatch accuracy while also modeling the joint distribution thereby reducing outlier sensitivity. We also present a preliminary outlier detection strategy based on Mahalanobis distance and discuss future improvements to this strategy.
Related Material