MY
M.D. Yang
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>
2 records found
1
Multi-modal 3D object detectors achieve state-of-the-art performance but remain notoriously brittle to asymmetric sensor degradation, such as when LiDAR point clouds become sparse in new environments. In this paper, we investigate unsupervised cross-modal adaptation to rescue a degraded sensor using an unaffected reference modality, without requiring target-domain labels. Using UniBEV on the nuScenes dataset, we simulate severe degradation by reducing LiDAR resolution from 32 to 8 beams. We systematically compare two leading adaptation paradigms anchored by the reliable camera stream: output-level camera pseudo-labeling and feature-level cross-modal mapping via a Bird's-Eye-View (BEV) Attention U-Net. Our experiments reveal a compelling insight: while feature mapping successfully aligns coarse spatial structures (improving LiDAR-only mAP by 5.6%), it fails to preserve fine-grained localization metrics. In contrast, simple confidence-filtered pseudo-labeling provides a significantly stronger recovery, yielding a 13.1% mAP improvement. Ultimately, our findings suggest that basic feature-level alignment may be insufficient to restore fine-grained 3D detection under severe spatial degradation, indicating that direct output-level supervision can be a more effective and reliable strategy for cross-modal adaptation in this regime.
...
Multi-modal 3D object detectors achieve state-of-the-art performance but remain notoriously brittle to asymmetric sensor degradation, such as when LiDAR point clouds become sparse in new environments. In this paper, we investigate unsupervised cross-modal adaptation to rescue a degraded sensor using an unaffected reference modality, without requiring target-domain labels. Using UniBEV on the nuScenes dataset, we simulate severe degradation by reducing LiDAR resolution from 32 to 8 beams. We systematically compare two leading adaptation paradigms anchored by the reliable camera stream: output-level camera pseudo-labeling and feature-level cross-modal mapping via a Bird's-Eye-View (BEV) Attention U-Net. Our experiments reveal a compelling insight: while feature mapping successfully aligns coarse spatial structures (improving LiDAR-only mAP by 5.6%), it fails to preserve fine-grained localization metrics. In contrast, simple confidence-filtered pseudo-labeling provides a significantly stronger recovery, yielding a 13.1% mAP improvement. Ultimately, our findings suggest that basic feature-level alignment may be insufficient to restore fine-grained 3D detection under severe spatial degradation, indicating that direct output-level supervision can be a more effective and reliable strategy for cross-modal adaptation in this regime.
Bachelor thesis
(2022)
-
M.T.K. Tran, M.D. Yang, W.C.A.M. van Nieuwburg, B.J.J. van Spreuwel, M.J. Tummers, J. Westerweel, G. Mulder
A 1:5 scaled rowing boat had been designed by a previous research group to determine the performance of rowing blades with different sizes and blade angles. In this research modifications were done to allow more control of the rowing motion. The aim of this study is to assess whether an angled oar blade can reach faster speeds with the same input power. This is done by collecting data with force and position sensors. With the collected data, the input power and speed of the rowing boat can be compared between several modified oar blades. The results of the tests in the towing tank show that it is very likely that the rowing boat can go faster with the same input power by adjusting the oar blade angle.
...
A 1:5 scaled rowing boat had been designed by a previous research group to determine the performance of rowing blades with different sizes and blade angles. In this research modifications were done to allow more control of the rowing motion. The aim of this study is to assess whether an angled oar blade can reach faster speeds with the same input power. This is done by collecting data with force and position sensors. With the collected data, the input power and speed of the rowing boat can be compared between several modified oar blades. The results of the tests in the towing tank show that it is very likely that the rowing boat can go faster with the same input power by adjusting the oar blade angle.