RailSafeNet: Visual Scene Understanding for Tram Safety

• URL: http://arxiv.org/abs/2509.12125v2
• Date: Tue, 16 Sep 2025 08:39:29 GMT
• Title: RailSafeNet: Visual Scene Understanding for Tram Safety
• Abstract summary: RailSafeNet is a real-time framework that fuses semantic segmentation, object detection and a distance assessor to highlight track intrusions.<n>Using only monocular video, the system identifies rails, localises nearby objects and classifies their risk by comparing projected distances with the standard 1435mm rail gauge.
• Score: 0.3437656066916039
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Tram-human interaction safety is an important challenge, given that trams frequently operate in densely populated areas, where collisions can range from minor injuries to fatal outcomes. This paper addresses the issue from the perspective of designing a solution leveraging digital image processing, deep learning, and artificial intelligence to improve the safety of pedestrians, drivers, cyclists, pets, and tram passengers. We present RailSafeNet, a real-time framework that fuses semantic segmentation, object detection and a rule-based Distance Assessor to highlight track intrusions. Using only monocular video, the system identifies rails, localises nearby objects and classifies their risk by comparing projected distances with the standard 1435mm rail gauge. Experiments on the diverse RailSem19 dataset show that a class-filtered SegFormer B3 model achieves 65% intersection-over-union (IoU), while a fine-tuned YOLOv8 attains 75.6% mean average precision (mAP) calculated at an intersection over union (IoU) threshold of 0.50. RailSafeNet therefore delivers accurate, annotation-light scene understanding that can warn drivers before dangerous situations escalate. Code available at https://github.com/oValach/RailSafeNet.

Related papers

• The Harsh Truth: Segment-Level Analysis of Harsh Driving Events in Milan Using Large-Scale Telematics, Street Networks, and Google Street View [3.6759575438610184]
  This study analyses harsh events across the urban road network of Milan.<n>Harsh acceleration and braking events are widely used as surrogate safety indicators, but have so far been studied only in comparatively small urban samples.
  arXiv  Detail & Related papers  (2026-05-29T18:46:55Z)
• AVOID: The Adverse Visual Conditions Dataset with Obstacles for Driving Scene Understanding [48.97660297411286]
  We introduce AVOID, a new dataset for real-time obstacle detection in a simulated environment.<n>AVOID consists of a large set of unexpected road obstacles located along each path captured under various weather and time conditions.<n>Each image is coupled with the corresponding semantic and depth maps, raw and semantic LiDAR data, and waypoints.
  arXiv  Detail & Related papers  (2025-12-29T05:34:26Z)
• Safety-Critical Learning for Long-Tail Events: The TUM Traffic Accident Dataset [6.420737230522813]
  We present the TUM Traffic Accident dataset, a collection of real-world highway accidents.<n>It contains ten sequences of vehicle crashes at high-speed driving with 294,924 labeled 2D and 93,012 labeled 3D boxes and track IDs within 48,144 labeled frames recorded from four roadside cameras and LiDARs at 10 Hz.<n>We propose Accid3nD, an accident detection model that combines a rule-based approach with a learning-based one.
  arXiv  Detail & Related papers  (2025-08-20T09:38:50Z)
• DAVE: Diverse Atomic Visual Elements Dataset with High Representation of Vulnerable Road Users in Complex and Unpredictable Environments [60.69159598130235]
  We present a new dataset, DAVE, designed for evaluating perception methods with high representation of Vulnerable Road Users (VRUs)<n>DAVE is a manually annotated dataset encompassing 16 diverse actor categories (spanning animals, humans, vehicles, etc.) and 16 action types (complex and rare cases like cut-ins, zigzag movement, U-turn, etc.)<n>Our experiments show that existing methods suffer degradation in performance when evaluated on DAVE, highlighting its benefit for future video recognition research.
  arXiv  Detail & Related papers  (2024-12-28T06:13:44Z)
• TrafficLoc: Localizing Traffic Surveillance Cameras in 3D Scenes [49.43995864524434]
  We propose a novel image-to-point cloud registration (I2P) method, TrafficLoc, in a coarse-tofine matching fashion.<n>To overcome the lack of large-scale real-world intersection datasets, we first introduce Carla Intersection, a new simulated dataset with 75 urban and rural intersections in Carla.<n>Our TrafficLoc greatly improves the performance over the SOTA I2P methods (up to 86%) on Carla Intersection and generalizes well to real-world data.
  arXiv  Detail & Related papers  (2024-12-13T17:42:53Z)
• Neural Semantic Map-Learning for Autonomous Vehicles [85.8425492858912]
  We present a mapping system that fuses local submaps gathered from a fleet of vehicles at a central instance to produce a coherent map of the road environment. Our method jointly aligns and merges the noisy and incomplete local submaps using a scene-specific Neural Signed Distance Field. We leverage memory-efficient sparse feature-grids to scale to large areas and introduce a confidence score to model uncertainty in scene reconstruction.
  arXiv  Detail & Related papers  (2024-10-10T10:10:03Z)
• Intelligent Railroad Grade Crossing: Leveraging Semantic Segmentation and Object Detection for Enhanced Safety [0.11999555634662631]
  Crashes and delays at Railroad Highway Grade Crossings (RHGC) pose significant safety concerns for the U.S. Federal Railroad Administration (FRA) This study introduces an intelligent system that leverages machine learning and computer vision techniques to enhance safety at RHGC.
  arXiv  Detail & Related papers  (2024-03-17T02:15:15Z)
• OpenLane-V2: A Topology Reasoning Benchmark for Unified 3D HD Mapping [84.65114565766596]
  We present OpenLane-V2, the first dataset on topology reasoning for traffic scene structure. OpenLane-V2 consists of 2,000 annotated road scenes that describe traffic elements and their correlation to the lanes. We evaluate various state-of-the-art methods, and present their quantitative and qualitative results on OpenLane-V2 to indicate future avenues for investigating topology reasoning in traffic scenes.
  arXiv  Detail & Related papers  (2023-04-20T16:31:22Z)
• CabiNet: Scaling Neural Collision Detection for Object Rearrangement with Procedural Scene Generation [54.68738348071891]
  We first generate over 650K cluttered scenes - orders of magnitude more than prior work - in diverse everyday environments. We render synthetic partial point clouds from this data and use it to train our CabiNet model architecture. CabiNet is a collision model that accepts object and scene point clouds, captured from a single-view depth observation.
  arXiv  Detail & Related papers  (2023-04-18T21:09:55Z)
• Rail Detection: An Efficient Row-based Network and A New Benchmark [16.811407800603664]
  We present a real-world railway dataset, Rail-DB, with 7432 pairs of images and annotations. We present an efficient row-based rail detection method, Rail-Net, containing a lightweight convolutional backbone and an anchor classifier. We evaluate the Rail-Net on Rail-DB with extensive experiments, including cross-scene settings and network backbones ranging from ResNet to Vision Transformers.
  arXiv  Detail & Related papers  (2023-04-12T07:44:50Z)
• Infrastructure-based End-to-End Learning and Prevention of Driver Failure [68.0478623315416]
  FailureNet is a recurrent neural network trained end-to-end on trajectories of both nominal and reckless drivers in a scaled miniature city. It can accurately identify control failures, upstream perception errors, and speeding drivers, distinguishing them from nominal driving. Compared to speed or frequency-based predictors, FailureNet's recurrent neural network structure provides improved predictive power, yielding upwards of 84% accuracy when deployed on hardware.
  arXiv  Detail & Related papers  (2023-03-21T22:55:51Z)
• TPE-Net: Track Point Extraction and Association Network for Rail Path Proposal Generation [1.9181913148426697]
  We propose a convolutional encoder-decoder architecture called TPE-Net to extract left-right rail pixels from 3D images. It generates strong rail route hypotheses without reliance on camera parameters, 3D data, and geometrical constraints.
  arXiv  Detail & Related papers  (2023-02-11T22:49:06Z)
• Detecting 32 Pedestrian Attributes for Autonomous Vehicles [103.87351701138554]
  In this paper, we address the problem of jointly detecting pedestrians and recognizing 32 pedestrian attributes. We introduce a Multi-Task Learning (MTL) model relying on a composite field framework, which achieves both goals in an efficient way. We show competitive detection and attribute recognition results, as well as a more stable MTL training.
  arXiv  Detail & Related papers  (2020-12-04T15:10:12Z)
• Identifying safe intersection design through unsupervised feature extraction from satellite imagery [9.306618776574025]
  The World Health Organization has listed the design of safer intersections as a key intervention to reduce global road trauma. Approximately 900,000 satellite images were downloaded for all intersections in Australia and customized computer vision techniques emphasized the road infrastructure. A deep autoencoder extracted high-level features, including the intersection's type, size, shape, lane markings, and complexity.
  arXiv  Detail & Related papers  (2020-10-29T03:42:09Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.