JK
J.F.P. Kooij
info
Please Note
<p>This page displays the records of the person named above and is not linked to a unique person identifier. This record may need to be merged to a profile.</p>
52 records found
1
Accurate segmentation and analysis of individual trees from 3D point clouds is a crucial yet challenging task in urbanism and environmental studies. Most existing methods for tree instance segmentation suffer from either under- or over-segmentation errors, mainly due to the complex nature of the environments and the varying tree geometries. In this paper, we propose SATree, a novel structure-aware approach that directly identifies important tree structures, such as crowns and stems, from point clouds, enabling robust tree instance segmentation against tree overlaps and varying tree sizes. Our method leverages a multi-task learning framework that simultaneously performs (i) semantic segmentation to classify a point as crown, stem, or other; (ii) heatmap prediction to assign a heat value to each point based on 2D Gaussian kernels centered at tree stem locations; (iii) offset prediction to estimate point-wise offset vectors pointing to the instance centroid. Key to our approach is the stem localization module, where we fuse the semantic and heatmap predictions to reliably localize tree stems from the network outputs. After that, we utilize a graph-based shortest path algorithm to group individual tree points by integrating the learned offset embeddings. Extensive experiments on two public forestry datasets, TreeML and ForInstance, demonstrate that SATree consistently outperforms state-of-the-art methods in terms of AP, AP50, and AP25 scores, reducing significant under- or over-segmentation errors. Our research output supports downstream forestry inventory, 3D tree reconstruction, and fine-grained part segmentation of trees. Our source code of SATree is available at https://github.com/shenglandu/SATree.
...
Accurate segmentation and analysis of individual trees from 3D point clouds is a crucial yet challenging task in urbanism and environmental studies. Most existing methods for tree instance segmentation suffer from either under- or over-segmentation errors, mainly due to the complex nature of the environments and the varying tree geometries. In this paper, we propose SATree, a novel structure-aware approach that directly identifies important tree structures, such as crowns and stems, from point clouds, enabling robust tree instance segmentation against tree overlaps and varying tree sizes. Our method leverages a multi-task learning framework that simultaneously performs (i) semantic segmentation to classify a point as crown, stem, or other; (ii) heatmap prediction to assign a heat value to each point based on 2D Gaussian kernels centered at tree stem locations; (iii) offset prediction to estimate point-wise offset vectors pointing to the instance centroid. Key to our approach is the stem localization module, where we fuse the semantic and heatmap predictions to reliably localize tree stems from the network outputs. After that, we utilize a graph-based shortest path algorithm to group individual tree points by integrating the learned offset embeddings. Extensive experiments on two public forestry datasets, TreeML and ForInstance, demonstrate that SATree consistently outperforms state-of-the-art methods in terms of AP, AP50, and AP25 scores, reducing significant under- or over-segmentation errors. Our research output supports downstream forestry inventory, 3D tree reconstruction, and fine-grained part segmentation of trees. Our source code of SATree is available at https://github.com/shenglandu/SATree.
NeuSEditor
From Multi-View Images to Text-Guided Neural Surface Edits
Implicit surface representations are valued for their compactness and continuity, but they pose significant challenges for editing. Despite recent advancements, existing methods often fail to preserve identity and maintain geometric consistency during editing. To address these challenges, we present NeuSEditor, a novel method for text-guided editing of neural implicit surfaces derived from multi-view images. NeuSEditor introduces an identity-preserving architecture that efficiently separates scenes into foreground and background, enabling precise modifications without altering the scene-specific elements. Our geometry-aware distillation loss significantly enhances rendering and geometric quality. Our method simplifies the editing workflow by eliminating the need for continuous dataset updates and source prompting. NeuSEditor outperforms recent state-of-the-art methods, delivering superior quantitative and qualitative results. for visual results, visit: neuseditor.github.io
...
Implicit surface representations are valued for their compactness and continuity, but they pose significant challenges for editing. Despite recent advancements, existing methods often fail to preserve identity and maintain geometric consistency during editing. To address these challenges, we present NeuSEditor, a novel method for text-guided editing of neural implicit surfaces derived from multi-view images. NeuSEditor introduces an identity-preserving architecture that efficiently separates scenes into foreground and background, enabling precise modifications without altering the scene-specific elements. Our geometry-aware distillation loss significantly enhances rendering and geometric quality. Our method simplifies the editing workflow by eliminating the need for continuous dataset updates and source prompting. NeuSEditor outperforms recent state-of-the-art methods, delivering superior quantitative and qualitative results. for visual results, visit: neuseditor.github.io
Scene flow estimation aims to recover per-point motion from two adjacent LiDAR scans. However, in real-world applications such as autonomous driving, points rarely move independently of others, especially for nearby points belonging to the same object, which often share the same motion. Incorporating this locally rigid motion constraint has been a key challenge in self-supervised scene flow estimation, which is often addressed by post-processing or appending extra regularization. While these approaches are able to improve the rigidity of predicted flows, they lack an architectural inductive bias for local rigidity within the model structure, leading to suboptimal learning efficiency and inferior performance. In contrast, we enforce local rigidity with a lightweight add-on module in neural network design, enabling end-to-end learning. We design a discretized voting space that accommodates all possible translations and then identify the one shared by nearby points by differentiable voting. Additionally, to ensure computational efficiency, we operate on pillars rather than points and learn representative features for voting per pillar. We plug the Voting Module into popular model designs and evaluate its benefit on Argoverse 2 and Waymo datasets. We outperform baseline works with only marginal compute overhead. Code is available at https://github.com/tudelft-iv/VoteFlow.
...
Scene flow estimation aims to recover per-point motion from two adjacent LiDAR scans. However, in real-world applications such as autonomous driving, points rarely move independently of others, especially for nearby points belonging to the same object, which often share the same motion. Incorporating this locally rigid motion constraint has been a key challenge in self-supervised scene flow estimation, which is often addressed by post-processing or appending extra regularization. While these approaches are able to improve the rigidity of predicted flows, they lack an architectural inductive bias for local rigidity within the model structure, leading to suboptimal learning efficiency and inferior performance. In contrast, we enforce local rigidity with a lightweight add-on module in neural network design, enabling end-to-end learning. We design a discretized voting space that accommodates all possible translations and then identify the one shared by nearby points by differentiable voting. Additionally, to ensure computational efficiency, we operate on pillars rather than points and learn representative features for voting per pillar. We plug the Voting Module into popular model designs and evaluate its benefit on Argoverse 2 and Waymo datasets. We outperform baseline works with only marginal compute overhead. Code is available at https://github.com/tudelft-iv/VoteFlow.
Conference paper
(2025)
-
O. De Groot, A. Bertipaglia, F. Tajdari, S. Wang, Z. Xia, M. Zaffar, R. Ensing, M. Garzon, J. Alonso-Mora, H. Caesar, L. Ferranti, R. Happee, H. Boekema, J. F.P. Kooij, G. Papaioannou, B. Shyrokau, D. M. Gavrila, V. Jain, M. Kegl, V. Kotian, T. Lentsch, Y. Lin, C. Messiou, E. Schippers
We present a vehicle system capable of navigating safely and efficiently around Vulnerable Road Users (VRUs), such as pedestrians and cyclists. The system comprises key modules for environment perception, localization and mapping, motion planning, and control, integrated into a prototype vehicle. A key innovation is a motion planner based on Topology-driven Model Predictive Control (T-MPC). The guidance layer generates multiple trajectories in parallel, each representing a distinct strategy for obstacle avoidance or non-passing. The underlying trajectory optimization constrains the joint probability of collision with VRUs under generic uncertainties. To address extraordinary situations ('edge cases') that go beyond the autonomous capabilities - such as construction zones or encounters with emergency responders - the system includes an option for remote human operation, supported by visual and haptic guidance. In simulation, our motion planner outperforms three baseline approaches in terms of safety and efficiency. We also demonstrate the full system in prototype vehicle tests on a closed track, both in autonomous and remotely operated modes.
...
We present a vehicle system capable of navigating safely and efficiently around Vulnerable Road Users (VRUs), such as pedestrians and cyclists. The system comprises key modules for environment perception, localization and mapping, motion planning, and control, integrated into a prototype vehicle. A key innovation is a motion planner based on Topology-driven Model Predictive Control (T-MPC). The guidance layer generates multiple trajectories in parallel, each representing a distinct strategy for obstacle avoidance or non-passing. The underlying trajectory optimization constrains the joint probability of collision with VRUs under generic uncertainties. To address extraordinary situations ('edge cases') that go beyond the autonomous capabilities - such as construction zones or encounters with emergency responders - the system includes an option for remote human operation, supported by visual and haptic guidance. In simulation, our motion planner outperforms three baseline approaches in terms of safety and efficiency. We also demonstrate the full system in prototype vehicle tests on a closed track, both in autonomous and remotely operated modes.
Journal article
(2025)
-
Hidde J.H. Boekema, Emran Yasser Moustafa, Julian F.P. Kooij, Dariu M. Gavrila
This paper studies road user trajectory prediction in mixed traffic, i.e. where vehicles and Vulnerable Road Users (VRUs, i.e. pedestrians, cyclists and other riders) closely share a common road space. We investigate if typical prediction components (scene graph representation, scene encoding, waypoint prediction, motion dynamics) should be specific to each road user class. Using the recent VRU-heavy View-of-Delft Prediction (VoD-P) dataset, we study several directions to improve the performance of the state-of-the-art map-based prediction models (PGP, TNT) in urban settings. First, we consider the use of class-specific map representations. Second, we investigate if the weights of different components of the model should be shared or separated by class. Finally, we augment VoD-P training data with automatically extracted trajectories from the 360-degree LiDAR scans by the recording vehicle. This data is made publicly available. We find that pre-training the model on auto-labels and making it class-specific leads to a reduction of up to 22.2%, 20.0%, and 18.2% in minADE (K = 10 samples) for pedestrians, cyclists, and vehicles, respectively.
...
This paper studies road user trajectory prediction in mixed traffic, i.e. where vehicles and Vulnerable Road Users (VRUs, i.e. pedestrians, cyclists and other riders) closely share a common road space. We investigate if typical prediction components (scene graph representation, scene encoding, waypoint prediction, motion dynamics) should be specific to each road user class. Using the recent VRU-heavy View-of-Delft Prediction (VoD-P) dataset, we study several directions to improve the performance of the state-of-the-art map-based prediction models (PGP, TNT) in urban settings. First, we consider the use of class-specific map representations. Second, we investigate if the weights of different components of the model should be shared or separated by class. Finally, we augment VoD-P training data with automatically extracted trajectories from the 360-degree LiDAR scans by the recording vehicle. This data is made publicly available. We find that pre-training the model on auto-labels and making it class-specific leads to a reduction of up to 22.2%, 20.0%, and 18.2% in minADE (K = 10 samples) for pedestrians, cyclists, and vehicles, respectively.
Given a ground-level query image and a geo-referenced aerial image that covers the query’s local surroundings, fine-grained cross-view localization aims to estimate the location of the ground camera inside the aerial image. Recent works have focused on developing advanced networks trained with accurate ground truth (GT) locations of ground images. However, the trained models always suffer a performance drop when applied to images in a new target area that differs from training. In most deployment scenarios, acquiring fine GT, i.e. accurate GT locations, for target-area images to re-train the network can be expensive and sometimes infeasible. In contrast, collecting images with noisy GT with errors of tens of meters is often easy. Motivated by this, our paper focuses on improving the performance of a trained model in a new target area by leveraging only the target-area images without fine GT. We propose a weakly supervised learning approach based on knowledge self-distillation. This approach uses predictions from a pre-trained model as pseudo GT to supervise a copy of itself. Our approach includes a mode-based pseudo GT generation for reducing uncertainty in pseudo GT and an outlier filtering method to remove unreliable pseudo GT. Our approach is validated using two recent state-of-the-art models on two benchmarks. The results demonstrate that it consistently and considerably boosts the localization accuracy in the target area.
...
Given a ground-level query image and a geo-referenced aerial image that covers the query’s local surroundings, fine-grained cross-view localization aims to estimate the location of the ground camera inside the aerial image. Recent works have focused on developing advanced networks trained with accurate ground truth (GT) locations of ground images. However, the trained models always suffer a performance drop when applied to images in a new target area that differs from training. In most deployment scenarios, acquiring fine GT, i.e. accurate GT locations, for target-area images to re-train the network can be expensive and sometimes infeasible. In contrast, collecting images with noisy GT with errors of tens of meters is often easy. Motivated by this, our paper focuses on improving the performance of a trained model in a new target area by leveraging only the target-area images without fine GT. We propose a weakly supervised learning approach based on knowledge self-distillation. This approach uses predictions from a pre-trained model as pseudo GT to supervise a copy of itself. Our approach includes a mode-based pseudo GT generation for reducing uncertainty in pseudo GT and an outlier filtering method to remove unreliable pseudo GT. Our approach is validated using two recent state-of-the-art models on two benchmarks. The results demonstrate that it consistently and considerably boosts the localization accuracy in the target area.
Adaptive Cruise Control Utilizing Noisy Multi-Leader Measurements
A Learning-Based Approach
A substantial number of vehicles nowadays are equipped with adaptive cruise control (ACC), which adjusts the vehicle speed automatically. However, experiments have found that commercial ACC systems which only detect the direct leader amplify the propagating disturbances in the platoon. This can cause severe traffic congestion when the number of ACC-equipped vehicles increases. Therefore, an ACC system which also considers the second leader further downstream is required. Such a system enables the vehicle to achieve multi-anticipation and hence ensure better platoon stability. Nevertheless, measurements collected from the second leader may be comparatively inaccurate given the limitations of current state-of-the-art sensor technology. This study adopts deep reinforcement learning to develop ACC controllers that besides the input from the first leader exploits the additional information obtained from the second leader, albeit noisy. The simulation experiment demonstrates that even under the influence of noisy measurements, the multi-leader ACC platoon shows smaller disturbance and jerk amplitudes than the one-leader ACC platoon, indicating improved string stability and ride comfort. Practical takeaways are twofold: first, the proposed method can be used to further develop multi-leader ACC systems. Second, even noisy data from the second leader can help stabilize traffic, which makes such systems viable in practice.
...
A substantial number of vehicles nowadays are equipped with adaptive cruise control (ACC), which adjusts the vehicle speed automatically. However, experiments have found that commercial ACC systems which only detect the direct leader amplify the propagating disturbances in the platoon. This can cause severe traffic congestion when the number of ACC-equipped vehicles increases. Therefore, an ACC system which also considers the second leader further downstream is required. Such a system enables the vehicle to achieve multi-anticipation and hence ensure better platoon stability. Nevertheless, measurements collected from the second leader may be comparatively inaccurate given the limitations of current state-of-the-art sensor technology. This study adopts deep reinforcement learning to develop ACC controllers that besides the input from the first leader exploits the additional information obtained from the second leader, albeit noisy. The simulation experiment demonstrates that even under the influence of noisy measurements, the multi-leader ACC platoon shows smaller disturbance and jerk amplitudes than the one-leader ACC platoon, indicating improved string stability and ride comfort. Practical takeaways are twofold: first, the proposed method can be used to further develop multi-leader ACC systems. Second, even noisy data from the second leader can help stabilize traffic, which makes such systems viable in practice.
Journal article
(2024)
-
Hidde J.H. Boekema, Bruno K.W. Martens, Julian F.P. Kooij, Dariu M. Gavrila
This letter presents View-of-Delft Prediction, a new dataset for trajectory prediction, to address the lack of on-board trajectory datasets in urban mixed-traffic environments. View-of-Delft Prediction builds on the recently released urban View-of-Delft (VoD) dataset to make it suitable for trajectory prediction. Unique features of this dataset are the challenging road layouts of Delft, with many narrow roads and bridges, and the close proximity between vehicles and Vulnerable Road Users (VRUs). It contains a large proportion of VRUs, with 569 prediction instances for vehicles, 347 for cyclists, and 934 for pedestrians. We additionally provide high-definition map annotations for the VoD dataset to enable state-of-the-art prediction models to be used. We analyse two state-of-the-art trajectory prediction models, PGP and P2T, which originally were developed for vehicle-dominated traffic scenarios, to assess the strengths and weaknesses of current modelling approaches in mixed traffic settings with large numbers of VRUs. Our analysis shows that there is a significant domain gap between the vehicle-dominated nuScenes and VRU-dominated VoD Prediction datasets. The dataset is publicly released for non-commercial research purposes.
...
This letter presents View-of-Delft Prediction, a new dataset for trajectory prediction, to address the lack of on-board trajectory datasets in urban mixed-traffic environments. View-of-Delft Prediction builds on the recently released urban View-of-Delft (VoD) dataset to make it suitable for trajectory prediction. Unique features of this dataset are the challenging road layouts of Delft, with many narrow roads and bridges, and the close proximity between vehicles and Vulnerable Road Users (VRUs). It contains a large proportion of VRUs, with 569 prediction instances for vehicles, 347 for cyclists, and 934 for pedestrians. We additionally provide high-definition map annotations for the VoD dataset to enable state-of-the-art prediction models to be used. We analyse two state-of-the-art trajectory prediction models, PGP and P2T, which originally were developed for vehicle-dominated traffic scenarios, to assess the strengths and weaknesses of current modelling approaches in mixed traffic settings with large numbers of VRUs. Our analysis shows that there is a significant domain gap between the vehicle-dominated nuScenes and VRU-dominated VoD Prediction datasets. The dataset is publicly released for non-commercial research purposes.
MuVieCAST
Multi-View Consistent Artistic Style Transfer
We introduce MuVieCAST, a modular multi-view consistent style transfer network architecture that enables consistent style transfer between multiple viewpoints of the same scene. This network architecture supports both sparse and dense views, making it versatile enough to handle a wide range of multi-view image datasets. The approach consists of three modules that perform specific tasks related to style transfer, namely content preservation, image transformation, and multi-view consistency enforcement. We extensively evaluate our approach across multiple application domains including depth-map-based point cloud fusion, mesh reconstruction, and novel-view synthesis. Our experiments reveal that the proposed framework achieves an exceptional generation of stylized images, exhibiting consistent outcomes across perspectives. A user study focusing on novel-view synthesis further confirms these results, with approximately 68% of cases participants expressing a preference for our generated outputs compared to the recent state-of-the-art method. Our modular framework is extensible and can easily be integrated with various backbone architectures, making it a flexible solution for multi-view style transfer.
...
We introduce MuVieCAST, a modular multi-view consistent style transfer network architecture that enables consistent style transfer between multiple viewpoints of the same scene. This network architecture supports both sparse and dense views, making it versatile enough to handle a wide range of multi-view image datasets. The approach consists of three modules that perform specific tasks related to style transfer, namely content preservation, image transformation, and multi-view consistency enforcement. We extensively evaluate our approach across multiple application domains including depth-map-based point cloud fusion, mesh reconstruction, and novel-view synthesis. Our experiments reveal that the proposed framework achieves an exceptional generation of stylized images, exhibiting consistent outcomes across perspectives. A user study focusing on novel-view synthesis further confirms these results, with approximately 68% of cases participants expressing a preference for our generated outputs compared to the recent state-of-the-art method. Our modular framework is extensible and can easily be integrated with various backbone architectures, making it a flexible solution for multi-view style transfer.
See Further Than CFAR
A Data-Driven Radar Detector Trained by Lidar
Conference paper
(2024)
-
Ignacio Roldan, Andras Palffy, Julian F.P. Kooij, Dariu M. Gavrila, Francesco Fioranelli, Alexander Yarovoy
In this paper, we address the limitations of traditional constant false alarm rate (CFAR) target detectors in automotive radars, particularly in complex urban environments with multiple objects that appear as extended targets. We propose a data-driven radar target detector exploiting a highly efficient 2D CNN backbone inspired by the computer vision domain. Our approach is distinguished by a unique cross-sensor supervision pipeline, enabling it to learn exclusively from unlabeled synchronized radar and lidar data, thuseliminating the need for costly manual object annotations. Using a novel large-scale, real-life multi-sensor dataset recorded in various driving scenarios, we demonstrate that the proposed detector generates dense, lidar-like point clouds, achieving a lower Chamfer distance to the reference lidar point clouds than CFAR detectors. Overall, it significantly outperforms CFAR baselines detection accuracy.
...
In this paper, we address the limitations of traditional constant false alarm rate (CFAR) target detectors in automotive radars, particularly in complex urban environments with multiple objects that appear as extended targets. We propose a data-driven radar target detector exploiting a highly efficient 2D CNN backbone inspired by the computer vision domain. Our approach is distinguished by a unique cross-sensor supervision pipeline, enabling it to learn exclusively from unlabeled synchronized radar and lidar data, thuseliminating the need for costly manual object annotations. Using a novel large-scale, real-life multi-sensor dataset recorded in various driving scenarios, we demonstrate that the proposed detector generates dense, lidar-like point clouds, achieving a lower Chamfer distance to the reference lidar point clouds than CFAR detectors. Overall, it significantly outperforms CFAR baselines detection accuracy.
Multi-sensor object detection is an active research topic in automated driving, but the robustness of such detection models against missing sensor input (modality missing), e.g., due to a sudden sensor failure, is a critical problem which remains under-studied. In this work, we propose UniBEV, an end-to-end multi-modal 3D object detection framework designed for robustness against missing modalities: UniBEV can operate on LiDAR plus camera input, but also on LiDAR-only or camera-only input without retraining. To facilitate its detector head to handle different input combinations, UniBEV aims to create well-aligned Bird’s Eye View (BEV) feature maps from each available modality. Unlike prior BEV-based multi-modal detection methods, all sensor modalities follow a uniform approach to resample features from the original sensor coordinate systems to the BEV features. We furthermore investigate the robustness of various fusion strategies w.r.t. missing modalities: the commonly used feature concatenation, but also channel-wise averaging, and a generalization to weighted averaging termed Channel Normalized Weights. To validate its effectiveness, we compare UniBEV to state-of-the-art BEVFusion and MetaBEV on nuScenes over all sensor input combinations. In this setting, UniBEV achieves better performance than these baselines for all input combinations. An ablation study shows the robustness benefits of fusing by weighted averaging over regular concatenation, and of sharing queries between the BEV encoders of each modality. Our code is available at https://github.com/tudelft-iv/UniBEV.
...
Multi-sensor object detection is an active research topic in automated driving, but the robustness of such detection models against missing sensor input (modality missing), e.g., due to a sudden sensor failure, is a critical problem which remains under-studied. In this work, we propose UniBEV, an end-to-end multi-modal 3D object detection framework designed for robustness against missing modalities: UniBEV can operate on LiDAR plus camera input, but also on LiDAR-only or camera-only input without retraining. To facilitate its detector head to handle different input combinations, UniBEV aims to create well-aligned Bird’s Eye View (BEV) feature maps from each available modality. Unlike prior BEV-based multi-modal detection methods, all sensor modalities follow a uniform approach to resample features from the original sensor coordinate systems to the BEV features. We furthermore investigate the robustness of various fusion strategies w.r.t. missing modalities: the commonly used feature concatenation, but also channel-wise averaging, and a generalization to weighted averaging termed Channel Normalized Weights. To validate its effectiveness, we compare UniBEV to state-of-the-art BEVFusion and MetaBEV on nuScenes over all sensor input combinations. In this setting, UniBEV achieves better performance than these baselines for all input combinations. An ablation study shows the robustness benefits of fusing by weighted averaging over regular concatenation, and of sharing queries between the BEV encoders of each modality. Our code is available at https://github.com/tudelft-iv/UniBEV.
We propose a novel end-to-end method for cross-view pose estimation. Given a ground-level query image and an aerial image that covers the query's local neighborhood, the 3 Degrees-of-Freedom camera pose of the query is estimated by matching its image descriptor to descriptors of local regions within the aerial image. The orientation-aware descriptors are obtained by using a translationally equivariant convolutional ground image encoder and contrastive learning. The Localization Decoder produces a dense probability distribution in a coarse-to-fine manner with a novel Localization Matching Upsampling module. A smaller Orientation Decoder produces a vector field to condition the orientation estimate on the localization. Our method is validated on the VIGOR and KITTI datasets, where it surpasses the state-of-the-art baseline by 72% and 36% in median localization error for comparable orientation estimation accuracy. The predicted probability distribution can represent localization ambiguity, and enables rejecting possible erroneous predictions. Without re-training, the model can infer on ground images with different field of views and utilize orientation priors if available. On the Oxford RobotCar dataset, our method can reliably estimate the ego-vehicle's pose over time, achieving a median localization error under 1 meter and a median orientation error of around 1 degree at 14 FPS.
...
We propose a novel end-to-end method for cross-view pose estimation. Given a ground-level query image and an aerial image that covers the query's local neighborhood, the 3 Degrees-of-Freedom camera pose of the query is estimated by matching its image descriptor to descriptors of local regions within the aerial image. The orientation-aware descriptors are obtained by using a translationally equivariant convolutional ground image encoder and contrastive learning. The Localization Decoder produces a dense probability distribution in a coarse-to-fine manner with a novel Localization Matching Upsampling module. A smaller Orientation Decoder produces a vector field to condition the orientation estimate on the localization. Our method is validated on the VIGOR and KITTI datasets, where it surpasses the state-of-the-art baseline by 72% and 36% in median localization error for comparable orientation estimation accuracy. The predicted probability distribution can represent localization ambiguity, and enables rejecting possible erroneous predictions. Without re-training, the model can infer on ground images with different field of views and utilize orientation priors if available. On the Oxford RobotCar dataset, our method can reliably estimate the ego-vehicle's pose over time, achieving a median localization error under 1 meter and a median orientation error of around 1 degree at 14 FPS.
AI in automotive is sterk in opkomst. Nieuwe voertuigen zijn always connected en worden zelflerende machines. Zo pakt de techniek een steeds grotere rol in de bediening en besturing van onze voertuigen. Wat betekent dat voor de veiligheid? En verandert dat ook de rol van de RDW?
...
AI in automotive is sterk in opkomst. Nieuwe voertuigen zijn always connected en worden zelflerende machines. Zo pakt de techniek een steeds grotere rol in de bediening en besturing van onze voertuigen. Wat betekent dat voor de veiligheid? En verandert dat ook de rol van de RDW?
Journal article
(2024)
-
I. Roldan Montero, A. Palffy, J.F.P. Kooij, D. Gavrila, F. Fioranelli, Alexander Yarovoy
The detection of multiple extended targets in complex environments using high-resolution automotive radar is considered. A data-driven approach is proposed where unlabeled synchronized lidar data are used as ground truth to train a neural network (NN) with only radar data as input. To this end, the novel, large-scale, real-life, and multisensor RaDelft dataset has been recorded using a demonstrator vehicle in different locations in the city of Delft, The Netherlands. The dataset, as well as the documentation and example code, is publicly available for those researchers in the field of automotive radar or machine perception. The proposed data-driven detector can generate lidar-like point clouds (PCs) using only radar data from a high-resolution system, which preserves the shape and size of extended targets. The results are compared against conventional constant false alarm rate (CFAR) detectors as well as variations of the method to emulate the available approaches in the literature, using the probability of detection, the probability of false alarm, and the Chamfer distance (CD) as performance metrics. Moreover, an ablation study was carried out to assess the impact of Doppler and temporal information on detection performance. The proposed method outperforms different baselines in terms of CD, achieving a reduction of 77% against conventional CFAR detectors and 28% against the modified state-of-the-art deep learning (DL)-based approaches.
...
The detection of multiple extended targets in complex environments using high-resolution automotive radar is considered. A data-driven approach is proposed where unlabeled synchronized lidar data are used as ground truth to train a neural network (NN) with only radar data as input. To this end, the novel, large-scale, real-life, and multisensor RaDelft dataset has been recorded using a demonstrator vehicle in different locations in the city of Delft, The Netherlands. The dataset, as well as the documentation and example code, is publicly available for those researchers in the field of automotive radar or machine perception. The proposed data-driven detector can generate lidar-like point clouds (PCs) using only radar data from a high-resolution system, which preserves the shape and size of extended targets. The results are compared against conventional constant false alarm rate (CFAR) detectors as well as variations of the method to emulate the available approaches in the literature, using the probability of detection, the probability of false alarm, and the Chamfer distance (CD) as performance metrics. Moreover, an ablation study was carried out to assess the impact of Doppler and temporal information on detection performance. The proposed method outperforms different baselines in terms of CD, achieving a reduction of 77% against conventional CFAR detectors and 28% against the modified state-of-the-art deep learning (DL)-based approaches.
Visual place recognition (VPR) is an image-based localization method that estimates the camera location of a query image by retrieving the most similar reference image from a map of geo-tagged reference images. In this work, we look into two fundamental bottlenecks for its localization accuracy: 1) reference map sparseness and 2) viewpoint invariance. First, the reference images for VPR are only available at sparse poses in a map, which enforces an upper bound on the maximum achievable localization accuracy through VPR. We, therefore, propose Continuous Place-descriptor Regression (CoPR) to densify the map and improve localization accuracy. We study various interpolation and extrapolation models to regress additional VPR feature descriptors from only the existing references. Second, we compare different feature encoders and show that CoPR presents value for all of them. We evaluate our models on three existing public datasets and report on average around 30% improvement in VPR-based localization accuracy using CoPR, on top of the 15% increase by using a viewpoint-variant loss for the feature encoder. The complementary relation between CoPR and relative pose estimation is also discussed.
...
Visual place recognition (VPR) is an image-based localization method that estimates the camera location of a query image by retrieving the most similar reference image from a map of geo-tagged reference images. In this work, we look into two fundamental bottlenecks for its localization accuracy: 1) reference map sparseness and 2) viewpoint invariance. First, the reference images for VPR are only available at sparse poses in a map, which enforces an upper bound on the maximum achievable localization accuracy through VPR. We, therefore, propose Continuous Place-descriptor Regression (CoPR) to densify the map and improve localization accuracy. We study various interpolation and extrapolation models to regress additional VPR feature descriptors from only the existing references. Second, we compare different feature encoders and show that CoPR presents value for all of them. We evaluate our models on three existing public datasets and report on average around 30% improvement in VPR-based localization accuracy using CoPR, on top of the 15% increase by using a viewpoint-variant loss for the feature encoder. The complementary relation between CoPR and relative pose estimation is also discussed.
We propose an enhancement module called depth discontinuity learning (DDL) for learning-based multi-view stereo (MVS) methods. Traditional methods are known for their accuracy but struggle with completeness. While recent learning-based methods have improved completeness at the cost of accuracy, our DDL approach aims to improve accuracy while retaining completeness in the reconstruction process. To achieve this, we introduce the joint estimation of depth and boundary maps, where the boundary maps are explicitly utilized for further refinement of the depth maps. We validate our idea by integrating it into an existing learning-based MVS pipeline where the reconstruction depends on high-quality depth map estimation. Extensive experiments on various datasets, namely DTU, ETH3D, “Tanks and Temples”, and BlendedMVS, show that our method improves reconstruction quality compared to our baseline, Patchmatchnet. Our ablation study demonstrates that incorporating the proposed DDL significantly reduces the depth map error, for instance, by more than 30% on the DTU dataset, and leads to improved depth map quality in both smooth and boundary regions. Additionally, our qualitative analysis has shown that the reconstructed point cloud exhibits enhanced quality without any significant compromise on completeness. Finally, the experiments reveal that our proposed model and strategies exhibit strong generalization capabilities across the various datasets.
...
We propose an enhancement module called depth discontinuity learning (DDL) for learning-based multi-view stereo (MVS) methods. Traditional methods are known for their accuracy but struggle with completeness. While recent learning-based methods have improved completeness at the cost of accuracy, our DDL approach aims to improve accuracy while retaining completeness in the reconstruction process. To achieve this, we introduce the joint estimation of depth and boundary maps, where the boundary maps are explicitly utilized for further refinement of the depth maps. We validate our idea by integrating it into an existing learning-based MVS pipeline where the reconstruction depends on high-quality depth map estimation. Extensive experiments on various datasets, namely DTU, ETH3D, “Tanks and Temples”, and BlendedMVS, show that our method improves reconstruction quality compared to our baseline, Patchmatchnet. Our ablation study demonstrates that incorporating the proposed DDL significantly reduces the depth map error, for instance, by more than 30% on the DTU dataset, and leads to improved depth map quality in both smooth and boundary regions. Additionally, our qualitative analysis has shown that the reconstructed point cloud exhibits enhanced quality without any significant compromise on completeness. Finally, the experiments reveal that our proposed model and strategies exhibit strong generalization capabilities across the various datasets.
This work addresses cross-view camera pose estimation, i.e., determining the 3-Degrees-of-Freedom camera pose of a given ground-level image w.r.t. an aerial image of the local area. We propose SliceMatch, which consists of ground and aerial feature extractors, feature aggregators, and a pose predictor. The feature extractors extract dense features from the ground and aerial images. Given a set of candidate camera poses, the feature aggregators construct a single ground descriptor and a set of pose-dependent aerial descriptors. Notably, our novel aerial feature aggregator has a cross-view attention module for ground-view guided aerial feature selection and utilizes the geometric projection of the ground camera's viewing frustum on the aerial image to pool features. The efficient construction of aerial descriptors is achieved using precomputed masks. SliceMatch is trained using contrastive learning and pose estimation is formulated as a similarity comparison between the ground descriptor and the aerial descriptors. Compared to the state-of-the-art, SliceMatch achieves a 19% lower median localization error on the VIGOR benchmark using the same VGG16 backbone at 150 frames per second, and a 50% lower error when using a ResNet50 backbone.
...
This work addresses cross-view camera pose estimation, i.e., determining the 3-Degrees-of-Freedom camera pose of a given ground-level image w.r.t. an aerial image of the local area. We propose SliceMatch, which consists of ground and aerial feature extractors, feature aggregators, and a pose predictor. The feature extractors extract dense features from the ground and aerial images. Given a set of candidate camera poses, the feature aggregators construct a single ground descriptor and a set of pose-dependent aerial descriptors. Notably, our novel aerial feature aggregator has a cross-view attention module for ground-view guided aerial feature selection and utilizes the geometric projection of the ground camera's viewing frustum on the aerial image to pool features. The efficient construction of aerial descriptors is achieved using precomputed masks. SliceMatch is trained using contrastive learning and pose estimation is formulated as a similarity comparison between the ground descriptor and the aerial descriptors. Compared to the state-of-the-art, SliceMatch achieves a 19% lower median localization error on the VIGOR benchmark using the same VGG16 backbone at 150 frames per second, and a 50% lower error when using a ResNet50 backbone.
Early and accurate detection of crossing pedestrians is crucial in automated driving in order to perform timely emergency manoeuvres. However, this is a difficult task in urban scenarios where pedestrians are often occluded (not visible) behind objects, e.g., other parked vehicles. We propose an occlusion aware fusion of stereo camera and radar sensors to address scenarios with crossing pedestrians behind such parked vehicles. Our proposed method adapts both the expected rate and properties of detections in different areas according to the visibility of the sensors. In our experiments on a real-world dataset, we show that the proposed occlusion aware fusion of radar and stereo camera detects the crossing pedestrians on average 0.26 seconds earlier than using the camera alone, and 0.15 seconds earlier than fusing the sensors without occlusion information. Our dataset containing 501 relevant recordings of pedestrians behind vehicles will be publicly available on our website for non-commercial, scientific use.
...
Early and accurate detection of crossing pedestrians is crucial in automated driving in order to perform timely emergency manoeuvres. However, this is a difficult task in urban scenarios where pedestrians are often occluded (not visible) behind objects, e.g., other parked vehicles. We propose an occlusion aware fusion of stereo camera and radar sensors to address scenarios with crossing pedestrians behind such parked vehicles. Our proposed method adapts both the expected rate and properties of detections in different areas according to the visibility of the sensors. In our experiments on a real-world dataset, we show that the proposed occlusion aware fusion of radar and stereo camera detects the crossing pedestrians on average 0.26 seconds earlier than using the camera alone, and 0.15 seconds earlier than fusing the sensors without occlusion information. Our dataset containing 501 relevant recordings of pedestrians behind vehicles will be publicly available on our website for non-commercial, scientific use.
State-of-the-art stixel methods fuse dense stereo disparity and semantic class information, e.g. from a Convolutional Neural Network (CNN), into a compact representation of driveable space, obstacles and background. However, they do not explicitly differentiate instances within the same semantic class. We investigate several ways to augment single-frame stixels with instance information, which can be extracted by a CNN from the RGB image input. As a result, our novel Instance Stixels method efficiently computes stixels that account for boundaries of individual objects, and represents instances as grouped stixels that express connectivity. Experiments on the Cityscapes dataset demonstrate that including instance information into the stixel computation itself, rather than as a post-processing step, increases the segmentation performance (i.e. Intersection over Union and Average Precision). This holds especially for overlapping objects of the same class. Furthermore, we show the superiority of our approach in terms of segmentation performance and computational efficiency compared to combining the separate outputs of Semantic Stixels and a state-of-the-art pixel-level CNN. We achieve processing throughput of 28 frames per second on average for 8 pixel wide stixels on images from the Cityscapes dataset at 1792x784 pixels. Our Instance Stixels software is made freely available for non-commercial research purposes.
...
State-of-the-art stixel methods fuse dense stereo disparity and semantic class information, e.g. from a Convolutional Neural Network (CNN), into a compact representation of driveable space, obstacles and background. However, they do not explicitly differentiate instances within the same semantic class. We investigate several ways to augment single-frame stixels with instance information, which can be extracted by a CNN from the RGB image input. As a result, our novel Instance Stixels method efficiently computes stixels that account for boundaries of individual objects, and represents instances as grouped stixels that express connectivity. Experiments on the Cityscapes dataset demonstrate that including instance information into the stixel computation itself, rather than as a post-processing step, increases the segmentation performance (i.e. Intersection over Union and Average Precision). This holds especially for overlapping objects of the same class. Furthermore, we show the superiority of our approach in terms of segmentation performance and computational efficiency compared to combining the separate outputs of Semantic Stixels and a state-of-the-art pixel-level CNN. We achieve processing throughput of 28 frames per second on average for 8 pixel wide stixels on images from the Cityscapes dataset at 1792x784 pixels. Our Instance Stixels software is made freely available for non-commercial research purposes.
Conference paper
(2022)
-
Frank M. Hafner, Matthias Zeller, Mark Schutera, Jochen Abhau, J.F.P. Kooij
Customization of a convolutional neural network (CNN) to a specific compute platform involves finding an optimal pareto state between computational complexity of the CNN and resulting throughput in operations per second on the compute platform. However, existing inference performance benchmarks compare complete backbones that entail many differences between their CNN configurations, which do not provide insights in how fine-grade layer design choices affect this balance.We present BackboneAnalysis, a methodology for extracting structured insights into the trade-off for a chosen target compute platform. Within a one-factor-at-a-time analysis setup, CNN architectures are systematically varied and evaluated based on throughput and latency measurements irrespective of model accuracy. Thereby, we investigate the configuration factors input shape, batch size, kernel size and convolutional layer type.In our experiments, we deploy BackboneAnalysis on a Xavier iGPU and a Coral Edge TPU accelerator. The analysis reveals that the general assumption from optimal Roofline performance that higher operation density in CNNs leads to higher throughput does not always hold. These results highlight the importance for a neural network architect to be aware of platform-specific latency and throughput behavior in order to derive sensible configuration decisions for a custom CNN.
...
Customization of a convolutional neural network (CNN) to a specific compute platform involves finding an optimal pareto state between computational complexity of the CNN and resulting throughput in operations per second on the compute platform. However, existing inference performance benchmarks compare complete backbones that entail many differences between their CNN configurations, which do not provide insights in how fine-grade layer design choices affect this balance.We present BackboneAnalysis, a methodology for extracting structured insights into the trade-off for a chosen target compute platform. Within a one-factor-at-a-time analysis setup, CNN architectures are systematically varied and evaluated based on throughput and latency measurements irrespective of model accuracy. Thereby, we investigate the configuration factors input shape, batch size, kernel size and convolutional layer type.In our experiments, we deploy BackboneAnalysis on a Xavier iGPU and a Coral Edge TPU accelerator. The analysis reveals that the general assumption from optimal Roofline performance that higher operation density in CNNs leads to higher throughput does not always hold. These results highlight the importance for a neural network architect to be aware of platform-specific latency and throughput behavior in order to derive sensible configuration decisions for a custom CNN.