The Impact of 2D Segmentation Backbones on Point Cloud Predictions Using 4D Radar
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2025
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866915544383356928 |
|---|---|
| author | Muckelroy III, William Alsakabi, Mohammed Dolan, John Tonguz, Ozan |
| author_facet | Muckelroy III, William Alsakabi, Mohammed Dolan, John Tonguz, Ozan |
| contents | LiDAR's dense, sharp point cloud (PC) representations of the surrounding environment enable accurate perception and significantly improve road safety by offering greater scene awareness and understanding. However, LiDAR's high cost continues to restrict the broad adoption of high-level Autonomous Driving (AD) systems in commercially available vehicles. Prior research has shown progress towards circumventing the need for LiDAR by training a neural network, using LiDAR point clouds as ground truth (GT), to produce LiDAR-like 3D point clouds using only 4D Radars. One of the best examples is a neural network created to train a more efficient radar target detector with a modular 2D convolutional neural network (CNN) backbone and a temporal coherence network at its core that uses the RaDelft dataset for training (see arXiv:2406.04723). In this work, we investigate the impact of higher-capacity segmentation backbones on the quality of the produced point clouds. Our results show that while very high-capacity models may actually hurt performance, an optimal segmentation backbone can provide a 23.7% improvement over the state-of-the-art (SOTA). |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2509_19644 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | The Impact of 2D Segmentation Backbones on Point Cloud Predictions Using 4D Radar Muckelroy III, William Alsakabi, Mohammed Dolan, John Tonguz, Ozan Computer Vision and Pattern Recognition Robotics I.4.6; I.2.9; I.2.10 LiDAR's dense, sharp point cloud (PC) representations of the surrounding environment enable accurate perception and significantly improve road safety by offering greater scene awareness and understanding. However, LiDAR's high cost continues to restrict the broad adoption of high-level Autonomous Driving (AD) systems in commercially available vehicles. Prior research has shown progress towards circumventing the need for LiDAR by training a neural network, using LiDAR point clouds as ground truth (GT), to produce LiDAR-like 3D point clouds using only 4D Radars. One of the best examples is a neural network created to train a more efficient radar target detector with a modular 2D convolutional neural network (CNN) backbone and a temporal coherence network at its core that uses the RaDelft dataset for training (see arXiv:2406.04723). In this work, we investigate the impact of higher-capacity segmentation backbones on the quality of the produced point clouds. Our results show that while very high-capacity models may actually hurt performance, an optimal segmentation backbone can provide a 23.7% improvement over the state-of-the-art (SOTA). |
| title | The Impact of 2D Segmentation Backbones on Point Cloud Predictions Using 4D Radar |
| topic | Computer Vision and Pattern Recognition Robotics I.4.6; I.2.9; I.2.10 |
| url | https://arxiv.org/abs/2509.19644 |