GraphAD: Interaction Scene Graph for End-to-end Autonomous Driving
Authors: Yunpeng Zhang, Deheng Qian, Ding Li, Yifeng Pan, Yong Chen, Zhenbao Liang, Zhiyao Zhang, Yingzong Liu, Jianhui Mei, Maolei Fu, Yun Ye, Zhujin Liang, Yi Shan, Dalong Du
IJCAI 2025 | Venue PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | We evaluate the proposed method for end-to-end autonomous driving on the nu Scenes dataset. Compared with strong baselines, our method significantly outperforms in full-stack driving tasks. ... We evaluate our method on the nu Scenes dataset [Caesar et al., 2020]. Extensive ablation studies are conducted to demonstrate the effectiveness of our design choices. ... 4 Experiments Our experiments are conducted on the challenging nu Scenes dataset [Caesar et al., 2020] |
| Researcher Affiliation | Industry | Yunpeng Zhang1 , Deheng Qian1 , Ding Li1 , Yifeng Pan1 , Yong Chen2 , Zhenbao Liang2 , Zhiyao Zhang2 , Yingzong Liu2 , Jianhui Mei2 , Maolei Fu2 , Yun Ye1 , Zhujin Liang1 , Yi Shan1 and Dalong Du1 1Phi Gent Robotics 2Geely Automobile Research Institute EMAIL, EMAIL |
| Pseudocode | No | The paper describes its methodology in text and uses figures and equations, but it does not contain any explicitly labeled 'Pseudocode' or 'Algorithm' blocks, nor does it present any structured code-like procedures. |
| Open Source Code | No | The paper does not provide any explicit statements about releasing source code, a link to a code repository, or mention of code being available in supplementary materials for the methodology described. |
| Open Datasets | Yes | We evaluate the proposed method for end-to-end autonomous driving on the nu Scenes dataset. ... We evaluate our method on the nu Scenes dataset [Caesar et al., 2020]. ... [Caesar et al., 2020] Holger Caesar, Varun Bankiti, Alex H Lang, Sourabh Vora, Venice Erin Liong, Qiang Xu, Anush Krishnan, Yu Pan, Giancarlo Baldan, and Oscar Beijbom. nuscenes: A multimodal dataset for autonomous driving. In CVPR, 2020. |
| Dataset Splits | Yes | Our experiments are conducted on the challenging nu Scenes dataset [Caesar et al., 2020]... Graph AD achieves the state-of-the-art planning performance. ... on the nu Scenes validation set. ... The benchmark results for motion prediction on the nu Scenes validation set are summarized in Tab. 2. |
| Hardware Specification | No | The paper describes model architectures and parameters in the 'Implementation Details' section but does not specify any hardware components like GPU models, CPU types, or memory used for running the experiments. It only mentions input sizes and backbone networks. |
| Software Dependencies | No | The paper mentions architectural components like 'ResNet101-DCN' and 'ResNet50', and refers to methods from other papers ('BEVDepth', 'Uni AD', 'VAD'), but it does not provide specific software library names with version numbers (e.g., Python, PyTorch, TensorFlow, CUDA versions) that would be needed for replication. |
| Experiment Setup | Yes | For benchmark results, Graph AD adopts the input size of 640 and Res Net101-DCN [He et al., 2016] as the image backbone. The image neck generates feature maps with 512 channels and 16 downsampling. For image-to-BEV transformation, Graph AD uses the method in BEVDepth [Li et al., 2022a] to generate the BEV features with 80 channels. Four frames of BEV features are fused to create the spatiotemporal scene representation FBEV R256 200 200. The Track Former strictly follows the settings of Uni AD [Hu et al., 2023], while the Map Former uses 100 map queries and a six-layer transformer decoder. The Interaction Scene Graph stacks three iterative layers for motion prediction with six modalities. The number of neighbours is set to 24 for the Dynamic Scene Graph and 8 for the Static Scene Graph. For ego-status features in the planning head, we follow the preprocessing of CAN-bus information from VAD [Jiang et al., 2023]. For ablation studies, we adopt the input size of 256 704 and Res Net50 as the image backbone. |