ApolloScape is a large dataset consisting of over 140,000 video frames (73 street scene videos) from various locations in China under varying weather conditions. Pixel-wise semantic annotation of the recorded data is provided in 2D, with point-wise semantic annotation in 3D for 28 classes. In addition, the dataset contains lane marking annotations in 2D.
Source: A2D2: Audi Autonomous Driving Dataset
Image Source: https://arxiv.org/pdf/1803.06184.pdf
Variants: ApolloScape
This dataset is used in 4 benchmarks:
| Task | Model | Paper | Date |
|---|---|---|---|
| Semantic Segmentation | deeplabv3 | VREM-FL: Mobility-Aware Computation-Scheduling Co-Design for … | 2023-11-30 |
| Trajectory Prediction | rule-based | PRANK: motion Prediction based on … | 2020-10-22 |
| Image Inpainting | DVI | DVI: Depth Guided Video Inpainting … | 2020-07-17 |
| Motion Segmentation | MRGCN | Understanding Dynamic Scenes using Graph … | 2020-05-09 |
| Motion Segmentation | MRGCN-LSTM | Understanding Dynamic Scenes using Graph … | 2020-05-09 |
| Motion Segmentation | St-RNN | Understanding Dynamic Scenes using Graph … | 2020-05-09 |
| Motion Segmentation | Rule Based | Understanding Dynamic Scenes using Graph … | 2020-05-09 |
| Motion Segmentation | Rel-Att-GCN | Understanding Dynamic Scenes using Graph … | 2020-05-09 |
| Semantic Segmentation | ERFNet-IntRA-KD (ours) | Inter-Region Affinity Distillation for Road … | 2020-04-11 |
Recent papers with results on this dataset: