DIST-C-SEGAN: A Novel Distributed Dimension Reduction Neural Framework for Visual Data

Abstract

This paper proposes a distributed unsupervised dimensionality reduction network framework for visual data, referred to as DIST-C-SEGAN. It aims to achieve low-bandwidth, high-fidelity, scalable, and privacy-friendly unsupervised dimensionality reduction and compression of large-scale image data, laying a solid foundation for downstream tasks in resource-constrained distributed or edge computing environments. This method captures local spatial information through a lightweight convolutional network and explores deep feature information using a graph neural propagation structure that combines compression and activation techniques with multi-head self-attention mechanisms. At the same time, it introduces the Proximal Alignment Algorithm to generate low-dimensional embedding features, achieving end-to-end "near-lossless" preservation of local neighborhood structure and global manifold continuity. Experimental results show that DIST-C-SEGAN achieves an original variance retention rate of over 98% in all data tasks, demonstrating high reliability and strong continuity. Its structural evaluation metrics are significantly superior to those of classic dimensionality reduction techniques, such as Principal Component Analysis (PCA), Kernel Principal Component Analysis (KPCA), Autoencoder (AE), Variational Autoencoder (VAE), UMAP, and t-SNE. Sensitivity analysis determines the optimal range of key hyperparameters and confirms their robustness to random parameter initialization. Ablation experiments verify the necessity of multi-head self-attention mechanisms and geometric alignment, while downstream tasks further confirm their effectiveness. Moreover, DIST-C-SEGAN theoretically reduces communication costs and provides new solutions for high-fidelity and communication-efficient compression and recovery in distributed environments.