# Training a Distributed Acoustic Sensing Traffic Monitoring Network With Video Inputs

arXiv: https://arxiv.org/abs/2412.12743 · HTML: https://mlsanity.com/paper/2412.12743/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 7 of its 58 references that are in the map.
- First posted on arXiv 17 Dec 2024.

## Builds on (most cited first, 7 of 7)

- [U-Net: Convolutional Networks for Biomedical Image Segmentation](https://mlsanity.com/paper/1505.04597/) (2015)
- [Image-to-Image Translation with Conditional Adversarial Networks](https://mlsanity.com/paper/1611.07004/) (2016)
- [A survey of loss functions for semantic segmentation](https://mlsanity.com/paper/2006.14822/) (2020)
- [A Survey on Hough Transform, Theory, Techniques and Applications](https://mlsanity.com/paper/1502.02160/) (2015)
- [Spatial Deep Deconvolution U-Net for Traffic Analyses with Distributed Acoustic Sensing](https://mlsanity.com/paper/2212.03936/) (2022)
- [City-scale dark fiber DAS measurements of infrastructure use during the COVID-19 pandemic](https://mlsanity.com/paper/2005.04861/) (2020)
- [Near-surface Characterization Using a Roadside Distributed Acoustic Sensing Array](https://mlsanity.com/paper/2006.01360/) (2020)

## Abstract (arXiv)

Distributed Acoustic Sensing (DAS) has emerged as a promising tool for real-time traffic monitoring in densely populated areas. In this paper, we present a novel concept that integrates DAS data with co-located visual information. We use YOLO-derived vehicle location and classification from camera inputs as labeled data to train a detection and classification neural network utilizing DAS data only. Our model achieves a performance exceeding 94% for detection and classification, and about 1.2% false alarm rate. We illustrate the model's application in monitoring traffic over a week, yielding statistical insights that could benefit future smart city developments. Our approach highlights the potential of combining fiber-optic sensors with visual information, focusing on practicality and scalability, protecting privacy, and minimizing infrastructure costs. To encourage future research, we share our dataset.