LWSIS: LiDAR-Guided Weakly Supervised Instance Segmentation for Autonomous Driving
Authors: Xiang Li, Junbo Yin, Botian Shi, Yikang Li, Ruigang Yang, Jianbing Shen
AAAI 2023 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Extensive experiments on the nu Ins Seg, as well as the large-scale Waymo, show that LWSIS can substantially improve existing weakly supervised segmentation models by only involving 3D data during training. |
| Researcher Affiliation | Collaboration | 1School of Computer Science, Beijing Institute of Technology 2Shanghai AI Laboratory 3Inceptio 4SKL-IOTSC, CIS, University of Macau |
| Pseudocode | No | The paper describes the modules (PLA and GCR) in detail using text and diagrams, but it does not provide structured pseudocode or algorithm blocks. |
| Open Source Code | Yes | The code and dataset are available at https://github.com/Serenos/LWSIS. |
| Open Datasets | Yes | Extensive experiments on the nu Ins Seg, as well as the large-scale Waymo, show that LWSIS can substantially improve existing weakly supervised segmentation models by only involving 3D data during training. ... The code and dataset are available at https://github.com/Serenos/LWSIS. ... nu Ins Seg is the first dataset that jointly contains Li DAR point cloud, RGB images, 3D bounding box, 2D bounding box and manually annotated instance mask that is consistent with the 2D and 3D box annotations. |
| Dataset Splits | Yes | The training set contains 789, 193 instance mask annotations aligned to 3D bounding box annotations over 168,780 images and the validation set has 157, 879 mask annotations over 36,114 images. |
| Hardware Specification | Yes | We train the model for 90K iterations with batch size 16 on 4 NVIDIA Tesla V100 GPUs |
| Software Dependencies | No | The paper mentions using Mask R-CNN and Cond Inst implemented with their official codebases, and ResNet backbones pretrained on ImageNet, but it does not specify exact version numbers for software libraries, frameworks, or programming languages (e.g., PyTorch version, Python version, CUDA version). |
| Experiment Setup | Yes | We train the model for 90K iterations with batch size 16 on 4 NVIDIA Tesla V100 GPUs |