Segment scene in new domain by internal representation by Gaussian Mixture Model with mas3-continual
Segment scene in new domain by internal representation by Gaussian Mixture Model with mas3-continual
Online Continual Domain Adaptation for Semantic Image Segmentation Using Internal Representations
arXiv paper abstract https://arxiv.org/abs/2401.01035
arXiv PDF paper https://arxiv.org/pdf/2401.01035.pdf
Semantic segmentation models trained on annotated data fail to generalize ... when the input data distribution changes
... Unsupervised domain adaptation (UDA) attempts to address a similar problem when there is target domain with no annotated data points through transferring knowledge from a source domain with annotated data.
... develop an online UDA algorithm for semantic segmentation of images that improves model generalization on unannotated domains in scenarios where source data access is restricted during adaptation.
... perform model adaptation is by minimizing the distributional distance between the source latent features and the target features in a shared embedding space ... solution promotes a shared domain-agnostic latent feature space between the two domains, which allows for classifier generalization on the target dataset.
To alleviate the need of access to source samples during adaptation, ... approximate the source latent feature distribution via an appropriate surrogate distribution, in this case a Gassian mixture model (GMM).
... evaluate ... approach on well established semantic segmentation datasets and demonstrate it compares favorably against state-of-the-art (SOTA) UDA semantic segmentation methods.
Please like and share this post if you enjoyed it using the buttons at the bottom!
Stay up to date. Subscribe to my posts https://morrislee1234.wixsite.com/website/contact
Web site with my other posts by category https://morrislee1234.wixsite.com/website
Comments