Multi-View In-Cabin Monitoring System for Public Transport Vehicles

2026-06-10 · Source: Artificial Intelligence · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Robotics & Autonomous Systems · Depth: Expert, quick

Summary

A new multi-view in-cabin monitoring dataset has been introduced for public transportation, specifically designed for a digitalized and partly automated German city bus. This dataset comprises 9,136 synchronized samples, featuring RGB and depth images from four inward-facing cameras, alongside data from a rotating LiDAR covering the vehicle's interior. It includes comprehensive annotations and is supported by a calibration and pseudo-labeling pipeline that generates 3D human pose estimates and oriented 3D bounding boxes for occupants. The release also offers a nuScenes-format conversion and benchmarks for representative multi-view 3D detection models, such as Lift-Splat-Shoot and BEVFusion. This resource facilitates comparative evaluation and small-scale training of multi-view in-cabin perception models, with all data and tools publicly available on GitHub.

Key takeaway

For computer vision engineers developing in-cabin monitoring systems for public transport or autonomous vehicles, this new dataset offers a critical resource. You should explore integrating this multi-modal data, including RGB, depth, and LiDAR, to enhance the robustness of your 3D detection and pose estimation models. Utilizing the provided nuScenes-format conversion and benchmarks can accelerate your model evaluation and small-scale training efforts, potentially improving occupant safety and service quality.

Key insights

A new multi-modal dataset and tools enable advanced in-cabin monitoring for public transport, supporting 3D perception model development.

Method

A calibration and pseudo-labeling pipeline generates 3D human pose estimates and oriented 3D bounding boxes from synchronized multi-view RGB-D and LiDAR data.

In practice

• Evaluate multi-view 3D detection models.
• Train in-cabin perception models.
• Convert data to nuScenes format.

Topics

• In-Cabin Monitoring
• Multi-View Perception
• 3D Object Detection
• Public Transport Safety
• LiDAR Data
• nuScenes Dataset

Code references

• EvgenyGorelik/multiview_incabin_dataset

Best for: Research Scientist, AI Scientist, Computer Vision Engineer, Robotics Engineer

Related on AIssential

• MOTOR: A Multimodal Dataset for Two-Wheeler Rider Behavior Understanding — The MOTOR dataset enables advanced two-wheeler rider behavior analysis through its multimodal, multi-view data and comprehensive annotations.
• Integrating Object Detection, LiDAR-Enhanced Depth Estimation, and Segmentation Models for Railway Environments — A modular framework integrates neural networks and LiDAR for railway obstacle detection and distance estimation.
• FRED: A Multi-Modal Autonomous Driving Dataset for Flooded Road Environments — FRED is the first multi-modal dataset for autonomous driving in flooded environments, enabling robust water hazard detection.
• DarkDriving: A Real-World Day and Night Aligned Dataset for Autonomous Driving in the Dark Environment — DarkDriving offers the first real-world dataset with precisely aligned day-night image pairs for autonomous driving.
• Comparative Analysis of Military Detection Using Drone Imagery Across Multiple Visual Spectrums — Diverse visual spectrum datasets significantly enhance drone-based military object detection robustness in varied environments.
• Why Rivian Uses BOTH Cameras and LiDAR — Multiple, diverse sensors provide robust, complementary environmental perception, overcoming individual sensor limitations.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Artificial Intelligence.