Person Detection and Tracking from an Overhead Crane LiDAR

This paper investigates person detection and tracking in an industrial indoor workspace using a LiDAR mounted on an overhead crane. The overhead viewpoint introduces a strong domain shift from common vehicle-centric LiDAR benchmarks, and limited availability of suitable public training data. Henceforth, we curate a site-specific overhead LiDAR dataset with 3D human bounding-box annotations and adapt selected candidate 3D detectors under a unified training and evaluation protocol. We further integrate lightweight tracking-by-detection using AB3DMOT and SimpleTrack to maintain person identities over time. Detection performance is reported with distance-sliced evaluation to quantify the practical operating envelope of the sensing setup. The best adapted detector configurations achieve average precision (AP) up to 0.84 within a 5.0 m horizontal radius, increasing to 0.97 at 1.0 m, with VoxelNeXt and SECOND emerging as the most reliable backbones across this range. The acquired results contribute in bridging the domain gap between standard driving datasets and overhead sensing for person detection and tracking. We also report latency measurements, highlighting practical real-time feasibility. Finally, we release our dataset and implementations in GitHub to support further research

Glass Segmentation with Fusion of Learned and General Visual Features

Glass surface segmentation from RGB images is a challenging task, since glass as a transparent material distinctly lacks visual characteristics. However, glass segmentation is critical for scene understanding and robotics, as transparent glass surfaces must be identified as solid material. This paper presents a novel architecture for glass segmentation, deploying a dual-backbone producing general visual features as well as task-specific learned visual features. General visual features are produced by a frozen DINOv3 vision foundation model, and the task-specific features are generated with a Swin model trained in a supervised manner. Resulting multi-scale feature representations are downsampled with residual Squeeze-and-Excitation Channel Reduction, and fed into a Mask2Former Decoder, producing the final segmentation masks. The architecture was evaluated on four commonly used glass segmentation datasets, achieving state-of-the-art results on several accuracy metrics. The model also has a competitive inference speed compared to the previous state-of-the-art method, and surpasses it when using a lighter DINOv3 backbone variant. The implementation source code and model weights are available at: https://github.com/ojalar/lgnet

Event-Only Drone Trajectory Forecasting with RPM-Modulated Kalman Filtering

Event cameras provide high-temporal-resolution visual sensing that is well suited for observing fast-moving aerial objects; however, their use for drone trajectory prediction remains limited. This work introduces an event-only drone forecasting method that exploits propeller-induced motion cues. Propeller rotational speed are extracted directly from raw event data and fused within an RPM-aware Kalman filtering framework. Evaluations on the FRED dataset show that the proposed method outperforms learning-based approaches and vanilla kalman filter in terms of average distance error and final distance error at 0.4s and 0.8s forecasting horizons. The results demonstrate robust and accurate short- and medium-horizon trajectory forecasting without reliance on RGB imagery or training data.

Automatic Labelling for Low-Light Pedestrian Detection

Pedestrian detection in RGB images is a key task in pedestrian safety, as the most common sensor in autonomous vehicles and advanced driver assistance systems is the RGB camera. A challenge in RGB pedestrian detection, that does not appear to have large public datasets, is low-light conditions. As a solution, in this research, we propose an automated infrared-RGB labeling pipeline. The proposed pipeline consists of 1) Infrared detection, where a fine-tuned model for infrared pedestrian detection is used 2) Label transfer process from the infrared detections to their RGB counterparts 3) Training object detection models using the generated labels for low-light RGB pedestrian detection. The research was performed using the KAIST dataset. For the evaluation, object detection models were trained on the generated autolabels and ground truth labels. When compared on a previously unseen image sequence, the results showed that the models trained on generated labels outperformed the ones trained on ground-truth labels in 6 out of 9 cases for the mAP@50 and mAP@50-95 metrics. The source code for this research is available at https://github.com/BouzoulasDimitrios/IR-RGB-Automated-LowLight-Pedestrian-Labeling

Label Correction for Road Segmentation Using Road-side Cameras

Reliable road segmentation in all weather conditions is critical for intelligent transportation applications, autonomous vehicles and advanced driver's assistance systems. For robust performance, all weather conditions should be included in the training data of deep learning-based perception models. However, collecting and annotating such a dataset requires extensive resources. In this paper, existing roadside camera infrastructure is utilized for collecting road data in varying weather conditions automatically. Additionally, a novel semi-automatic annotation method for roadside cameras is proposed. For each camera, only one frame is labeled manually and then the label is transferred to other frames of that camera feed. The small camera movements between frames are compensated using frequency domain image registration. The proposed method is validated with roadside camera data collected from 927 cameras across Finland over 4 month time period during winter. Training on the semi-automatically labeled data boosted the segmentation performance of several deep learning segmentation models. Testing was carried out on two different datasets to evaluate the robustness of the resulting models. These datasets were an in-domain roadside camera dataset and out-of-domain dataset captured with a vehicle on-board camera.

Trajectory-based Road Autolabeling with Lidar-Camera Fusion in Winter Conditions

Robust road segmentation in all road conditions is required for safe autonomous driving and advanced driver assistance systems. Supervised deep learning methods provide accurate road segmentation in the domain of their training data but cannot be trusted in out-of-distribution scenarios. Including the whole distribution in the trainset is challenging as each sample must be labeled by hand. Trajectory-based self-supervised methods offer a potential solution as they can learn from the traversed route without manual labels. However, existing trajectory-based methods use learning schemes that rely only on the camera or only on the lidar. In this paper, trajectory-based learning is implemented jointly with lidar and camera for increased performance. Our method outperforms recent standalone camera- and lidar-based methods when evaluated with a challenging winter driving dataset including countryside and suburb driving scenes. The source code is available at https://github.com/eerik98/lidar-camera-road-autolabeling.git

Road Surface Friction Estimation for Winter Conditions Utilising General Visual Features

In below freezing winter conditions, road surface friction can greatly vary based on the mixture of snow, ice, and water on the road. Friction between the road and vehicle tyres is a critical parameter defining vehicle dynamics, and therefore road surface friction information is essential to acquire for several intelligent transportation applications, such as safe control of automated vehicles or alerting drivers of slippery road conditions. This paper explores computer vision-based evaluation of road surface friction from roadside cameras. Previous studies have extensively investigated the application of convolutional neural networks for the task of evaluating the road surface condition from images. Here, we propose a hybrid deep learning architecture, WCamNet, consisting of a pretrained visual transformer model and convolutional blocks. The motivation of the architecture is to combine general visual features provided by the transformer model, as well as finetuned feature extraction properties of the convolutional blocks. To benchmark the approach, an extensive dataset was gathered from national Finnish road infrastructure network of roadside cameras and optical road surface friction sensors. Acquired results highlight that the proposed WCamNet outperforms previous approaches in the task of predicting the road surface friction from the roadside camera images.

DynaHull: Density-centric Dynamic Point Filtering in Point Clouds

In the field of indoor robotics, accurately navigating and mapping in dynamic environments using point clouds can be a challenging task due to the presence of dynamic points. These dynamic points are often represented by people in indoor environments, but in industrial settings with moving machinery, there can be various types of dynamic points. This study introduces DynaHull, a novel technique designed to enhance indoor mapping accuracy by effectively removing dynamic points from point clouds. DynaHull works by leveraging the observation that, over multiple scans, stationary points have a higher density compared to dynamic ones. Furthermore, DynaHull addresses mapping challenges related to unevenly distributed points by clustering the map into smaller sections. In each section, the density factor of each point is determined by dividing the number of neighbors by the volume these neighboring points occupy using a convex hull method. The algorithm removes the dynamic points using an adaptive threshold based on the point count of each cluster, thus reducing the false positives. The performance of DynaHull was compared to state-of-the-art techniques, such as ERASOR, Removert, OctoMap, and a baseline statistical outlier removal from Open3D, by comparing each method to the ground truth map created during a low activity period in which only a few dynamic points were present. The results indicated that DynaHull outperformed these techniques in various metrics, noticeably in the Earth Mover's Distance. This research contributes to indoor robotics by providing efficient methods for dynamic point removal, essential for accurate mapping and localization in dynamic environments.

TADAP: Trajectory-Aided Drivable area Auto-labeling with Pre-trained self-supervised features in winter driving conditions

Detection of the drivable area in all conditions is crucial for autonomous driving and advanced driver assistance systems. However, the amount of labeled data in adverse driving conditions is limited, especially in winter, and supervised methods generalize poorly to conditions outside the training distribution. For easy adaption to all conditions, the need for human annotation should be removed from the learning process. In this paper, Trajectory-Aided Drivable area Auto-labeling with Pre-trained self-supervised features (TADAP) is presented for automated annotation of the drivable area in winter driving conditions. A sample of the drivable area is extracted based on the trajectory estimate from the global navigation satellite system. Similarity with the sample area is determined based on pre-trained self-supervised visual features. Image areas similar to the sample area are considered to be drivable. These TADAP labels were evaluated with a novel winter-driving dataset, collected in varying driving scenes. A prediction model trained with the TADAP labels achieved a +9.6 improvement in intersection over union compared to the previous state-of-the-art of self-supervised drivable area detection.

Enhanced Winter Road Surface Condition Monitoring with Computer Vision

Winter conditions pose several challenges for automated driving applications. A key challenge during winter is accurate assessment of road surface condition, as its impact on friction is a critical parameter for safely and reliably controlling a vehicle. This paper proposes a deep learning regression model, SIWNet, capable of estimating road surface friction properties from camera images. SIWNet extends state of the art by including an uncertainty estimation mechanism in the architecture. This is achieved by including an additional head in the network, which estimates a prediction interval. The prediction interval head is trained with a maximum likelihood loss function. The model was trained and tested with the SeeingThroughFog dataset, which features corresponding road friction sensor readings and images from an instrumented vehicle. Acquired results highlight the functionality of the prediction interval estimation of SIWNet, while the network also achieved similar point estimate accuracy as the previous state of the art. Furthermore, the SIWNet architecture is several times more lightweight than the previously applied state-of-the-art model, resulting in more practical and efficient deployment.