Efficient Transceiver Design for Aerial Image Transmission and Large-scale Scene Reconstruction

2026-04-13

AI Summary

This work addresses the challenge of balancing wireless image transmission efficiency and reconstruction fidelity in large-scale 3D scene reconstruction for low-altitude intelligent networks, where existing methods struggle to reconcile pilot overhead with accuracy. The paper proposes an end-to-end deep learning-based transceiver architecture that, for the first time, integrates 3D Gaussian Splatting (3DGS) rendering as a task-driven objective directly into the communication system’s training process. By jointly optimizing the communication modules and the 3DGS rendering loss, the approach enables semantic-level efficient transmission tailored for 3D reconstruction. It supports sparse pilot signaling and demonstrates significant performance gains over current baselines on real aerial datasets, achieving higher-fidelity large-scale 3D reconstructions while substantially reducing pilot overhead.

Technology Category

• Computer Vision: 3D Computer Vision
• Planning, Routing, and Scheduling: Optimization of Spatio-temporal Systems
• Machine Learning: Deep Neural Architectures and Foundation Models

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs

Abstract

Large-scale three-dimensional (3D) scene reconstruction in low-altitude intelligent networks (LAIN) demands highly efficient wireless image transmission. However, existing schemes struggle to balance severe pilot overhead with the transmission accuracy required to maintain reconstruction fidelity. To strike a balance between efficiency and reliability, this paper proposes a novel deep learning-based end-to-end (E2E) transceiver design that integrates 3D Gaussian Splatting (3DGS) directly into the training process. By jointly optimizing the communication modules via the combined 3DGS rendering loss, our approach explicitly improves scene recovery quality. Furthermore, this task-driven framework enables the use of a sparse pilot scheme, significantly reducing transmission overhead while maintaining robust image recovery under low-altitude channel conditions. Extensive experiments on real-world aerial image datasets demonstrate that the proposed E2E design significantly outperforms existing baselines, delivering superior transmission performance and accurate 3D scene reconstructions.

Problem

• 3D scene reconstruction
• wireless image transmission
• pilot overhead
• transmission accuracy
• low-altitude intelligent networks

Innovation

• end-to-end transceiver
• 3D Gaussian Splatting
• task-driven communication
• sparse pilot
• low-altitude intelligent networks