This report summarizes the work submitted by team arnav007 (submission ID 1054222) for the UrbanTwin Sim2Real LiDAR Challenge (V2X-Real track) at the IEEE BigData Cup 2026. The primary objective is to train 3D detectors for vehicles, pedestrians, and trucks using only synthetic roadside LiDAR data, and to generate synthetic scans that are realistic enough to be indistinguishable from real ones. A CARLA based pipeline is developed and calibrated to the fixed sensor site, accurately modeling the range image, including no-return beams, based on unlabelled real development scans. A combined inference detector, using multiple 3D detectors (VoxelNeXt, PV-RCNN, and SECOND) trained exclusively on synthetic data, is also proposed. Each scan is aligned to the synthetic ground plane and processed through the detectors at multiple ground heights. The resulting bounding boxes are fused using bird's-eye view (BEV) clustering, followed by class-specific post-processing to refine heading, box size, and filter low-confidence detections. The final submission achieved a combined score of 0.8947. The ablation study shows that per-scan ground leveling more than doubled the real-scan mAP, while fusing boxes across different heights and models further improved performance.
Figures & tables
Dataset
Frames
Use
Canonical synthetic scenes
10,800
Training
Site-calibrated twin scenes
5,000
Training
Small-truck twin scenes
3,000
Training
Merged synthetic set a
18,800
Training
Size-varied pedestrian twin scenes
5,000
Training b
Long-low truck twin scenes
4,000
Training b
TABLE I: Data used for detector training and evaluation
Fig. 1: Overview of the proposed pipeline, from site calibration and synthetic data generation to detector training and evaluation on real roadside LiDAR scans.
SECOND
VoxelNeXt
PV-RCNN
Voxel size (m)
0.2×0.2×0.25
0.1×0.1×0.2
0.05×0.05×0.25
Learning rate
0.003
0.001
0.002
Batch size
4
2–4
4
Epochs
12
12–20 a
4–6
TABLE II: Hyperparameter settings of the three detector families
Entry
Combined Score
Vehicle
Pedestrian
Truck
mAP 0.7
Real
Ours: fusion, calibration
0.7997
63.76
31.30
52.92
0.2104
0.9423
+ truck rule, ped. TTA
0.8475
69.07
35.99
59.66
0.2329
0.9423
+ heading, de-dup.
0.8600
74.30
35.20
59.58
0.2104
0.9423
+ ped. hedging
0.8732
74.25
39.66
59.79
0.2166
0.9423
+ PV-RCNN, scans
0.8853
75.49
41.51
59.75
0.2288
0.9507
+ ped. specialists, per-scan levelling
0.8875
75.59
43.68
59.06
0.2628
0.9449
TABLE III: Submissions and leaderboard entries (AP in %, IoU 0.5, R40; leaderboard rows are public entries as of Nov. 15 at 9:00 PM EST
Component
Metric
150 frames
350 frames
Per-scan ground levelling a
mAP
15.8 → 37.2
Multi-height + multi-model fusion a
mAP
49.3 → 56.3
Averaged vehicle heading
Vehicle AP
75.4 → 80.0
75.4 → 79.7
Truck de-duplication
Genuine-truck AP
66.2 → 71.0
63.7 → 68.4
Pedestrian hedging
Pedestrian AP
38.5 → 42.4
37.1 → 41.3
PV-RCNN voters
Vehicle AP
82.5 → 83.5
82.4 → 83.5
TABLE IV: Effect of each component on the development sets (150 / 350 frames; AP in %, realism on the 0–1 scale)
Lidar-based 3D object detection locates and classifies objects in point clouds, supporting perception for autonomous vehicles, robots, and other mobile systems.
We present our solution to the LUMPI track of the UCF UrbanTwin Sim2Real LiDAR Challenge at the 6th DriveX Workshop, ECCV 2026. The detector must be trained only on synthetic data and is evaluated on 50 held-out real LiDAR frames; a separate 50-frame synthetic submission is evaluated for point-cloud realism. Our method addresses the Sim2Real gap at three levels. First, we align synthetic scans to the 50k-point test density and build a 30k-record training pool using UT-LUMPI geometry, RangeLDM-based sampling diversification, rare-class copy-paste, and pedestrian-oriented augmentation. Second, complementary DSVT detectors and Car/Bus PointPillars specialists are trained under the same synthetic-only constraint. Third, predictions are integrated by class-aware routing, asymmetric agreement fusion, constrained residual-recall supplementation, class-coverage auditing, and selective box-size calibration. The realism branch is optimized independently with radial-density matching, weak affine calibration, and calibrated set mixing. The final submission obtains a Combined Score of 0.4692, a Detection Score of 0.1797, a Realism Score of 0.9035, and 3D [email protected] of 0.1258.
Bridging the simulation-to-reality gap in roadside LiDAR requires addressing several coupled discrepancies, including scene geometry, sampling density, return patterns, and pedestrian scale. This report presents a multi-source collaborative training and class-aware fusion framework for Sim2Real 3D detection. The method organizes digital-twin scans, diffusion-redrawn scans, density-stabilized scans, and pedestrian morphology-aligned samples into a unified training pool with complementary roles. Within a common DSVT detection formulation, source-specialized expert branches preserve those roles while optimizing for the same detection objective. At inference, a predefined class-aware fusion pathway integrates geometry-stable and calibration-aware branches for vehicles, sampling-complementary branches for trucks, and morphology-consistent evidence for pedestrians. A label-free point-cloud center blend then refines geometric localization. On the UrbanTwin V2X-Real hidden test set, the unified system achieves a combined score of 0.7421, with 3D [email protected] of 0.4518 and a realism score of 0.8871. The results indicate that a stable, interpretable collaboration among data sources is more valuable than unconstrained aggregation of model outputs.
We describe our approach to the LUMPI track of the UrbanTwin Sim2Real LiDAR Challenge, in which a 3D object detector must be trained exclusively on synthetic data and evaluated on unseen real roadside LiDAR, while the synthetic point clouds are simultaneously scored for distributional realism. Our final entry scores 0.4793 combined, from a detection mAP of 0.1333 and a realism score of 0.9126. Our contribution is not a new architecture. Working from a PointPillars baseline, we obtained most of our gains from three sources. First, a diagnosis that small object failure in roadside LiDAR is a localisation problem rather than a detection problem, which led to a parameter free geometric refinement worth 86 percent relative on pedestrian AP. Second, an audit of our own post processing which revealed that a shipped rule was destroying valid predictions, the repair of which produced our single largest detection gain at 335 percent relative on the Van class. Third, treating the realism metrics analytically rather than as a black box, which showed that one of the four components is a closed form function of nine summary statistics. We also report an experimental methodology that we believe is the more transferable result. Because mean AP decomposes into independent per class terms, a single submission can carry several mutually controlled experiments, and a submission assembled entirely from previously measured components has a score that is known before it is sent. We predicted our final score to four decimal places prior to submission. We report negative results at equal length, and give a full account of data provenance and compliance in Section 8, including a full account of every mechanism contributing to the submitted score.