LidarMultiNet: Towards a Unified Multi-Task Network for LiDAR Perception
Authors: Dongqiangzi Ye, Zixiang Zhou, Weijia Chen, Yufei Xie, Yu Wang, Panqu Wang, Hassan Foroosh
AAAI 2023 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Lidar Multi Net is extensively tested on both Waymo Open Dataset and nu Scenes dataset, demonstrating for the first time that major Li DAR perception tasks can be unified in a single strong network that is trained end-to-end and achieves state-of-the-art performance. Notably, Lidar Multi Net reaches the official 1st place in the Waymo Open Dataset 3D semantic segmentation challenge 2022 with the highest m Io U and the best accuracy for most of the 22 classes on the test set, using only Li DAR points as input. It also sets the new state-of-the-art for a single model on the Waymo 3D object detection benchmark and three nu Scenes benchmarks. |
| Researcher Affiliation | Collaboration | Dongqiangzi Ye1*, Zixiang Zhou1,2* , Weijia Chen1*, Yufei Xie1*, Yu Wang1, Panqu Wang1, Hassan Foroosh2 1 Tu Simple 2 University of Central Florida |
| Pseudocode | No | The paper includes architectural diagrams (Figure 2, Figure 3, Figure 4) but does not provide any pseudocode or explicitly labeled algorithm blocks. |
| Open Source Code | No | The paper does not explicitly state that its source code is open-source or provide a link to a code repository for the methodology described. It only provides a link to its arXiv version. |
| Open Datasets | Yes | Lidar Multi Net is extensively tested on both Waymo Open Dataset and nu Scenes dataset, demonstrating for the first time that major Li DAR perception tasks can be unified in a single strong network that is trained end-to-end and achieves state-of-the-art performance. We perform extensive tests of the proposed Lidar Multi Net on five major benchmarks of the large-scale Waymo Open Dataset (Sun et al. 2020) (3D Object Detection and 3D Semantic Segmentation) and nu Scenes dataset (Caesar et al. 2020; Fong et al. 2022) (Detection, Li DAR Segmentation, and Panoptic Segmentation). |
| Dataset Splits | Yes | Waymo Open Dataset (WOD) contains 1150 sequences in total, split into 798 in the training set, 202 in the validation set, and 150 in the test set. Nu Scenes contains 1000 scenes with 20 seconds duration each, split into 700 in the training set, 150 in the validation set, and 150 in the test set. |
| Hardware Specification | Yes | We use a batch size of 2 on each of the 8 A100 GPUs. |
| Software Dependencies | No | The paper mentions the use of the AdamW optimizer but does not specify version numbers for any software dependencies, libraries, or programming languages used in the implementation. |
| Experiment Setup | Yes | We train the models using Adam W (Loshchilov and Hutter 2017) optimizer with one-cycle learning rate policy, with a max learning rate of 3e-3, a weight decay of 0.01, and a momentum ranging from 0.85 to 0.95. We use a batch size of 2 on each of the 8 A100 GPUs. For the one-stage model, we train the models from scratch for 20 epochs. For the two-stage model, we freeze the 1st stage and finetune the 2nd stage for 6 epochs. We apply separate detection heads in the detection branch for different categories. During training, we employ data augmentation which includes standard random flipping, and global scaling, rotation and translation. We also adopt the ground-truth sampling (Yan, Mao, and Li 2018) with the fade strategy (Wang et al. 2021). |