Analog noise adaptation improves depth from drone video transmission
AnalogDepth: Multi-view Geometry from FPV drones under Analog Video Transmission
Computer Vision and Pattern RecognitionRobotics
Summary
Low-cost drones often use old-style analog video signals that look blurry and noisy in special ways. This weird noise makes it hard for existing AI to understand how far away things are in these videos. The authors made a new method called AnalogDepth, which teaches the AI to handle this strange noise by learning from real noisy drone videos instead of fake noise. Their approach helps drones estimate depth more accurately during flight, which can improve 3D mapping and navigation.
What this means in practice
- •For fpv drone developers: Improve onboard depth and 3D mapping accuracy by adapting vision models to realistic analog video noise encountered in low-latency drone systems.
- •For robotics engineers: Enhance spatial perception for robots using analog video inputs where typical digital noise models fail, enabling better navigation and scene understanding.
Authors
André Amorim, Pedro F. Proença
Abstract
Analog video transmission (VTX) remains widespread in FPV drones due to low latency, weight and low cost. However analog VTX suffers from complex spatially structured image degradation which differ fundamentally from digital image corruption (e.g. AWGN) used in standard training augmentation. This work shows that this type of noise severely degrades the accuracy of Depth Anything 3 (DA3), a state-of-the-art feed forward visual geometry foundation model. To address this gap, we present AnalogDepth, a parameter-efficient training pipeline that adapts DA3 to analog FPV imagery using student-teacher knowledge distillation with Low-Rank Adaptation (LoRA) injected into the DINOv2 backbone. Rather than synthesizing noise analytically, we build a noise bank from static FPV recordings under diverse conditions and compare real-noise injection against PSD-matched Gaussian synthesis and AWGN as baselines. Experiments on six real FPV flight sequences across three indoor scenes show that training with our noise bank consistently reduces per-frame depth RMSE and 3D reconstruction Chamfer distance compared to the pretrained DA3 baseline and both Gaussian noise variants. These results demonstrate that replicating the spatial structure of real analog transmission noise is critical for effective adaptation.