Research map / 3D vision / 3D generation
H-V2X: A Large Scale Highway Dataset for BEV Perception
Chang Liu, MingXu zhu, Cong Ma
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Use early fusion (H-BEV with fisheye and vector map) instead of late fusion | BEV detection · H-V2X BEV detection | mAP3D|R40: 31.75 -> 35.49 | Improved |
| Use early fusion instead of late fusion | BEV detection · H-V2X BEV detection | FPS: 10 -> 6 | Hurt |
| Derive a single depth value per pixel from HDMap instead of learning discrete depth distributions | H-BEV (built on BEVDet) · H-V2X BEV detection | image feature map size / speed: feature map size reduced 500 times | Improved |
| Add global HDMap lane features to SocialGAN (HD-GAN) | SocialGAN · H-V2X trajectory prediction (5s) | ADE: 2.683 -> 2.457 | Improved |
| Add global HDMap lane features to SocialGAN (HD-GAN) | SocialGAN · H-V2X trajectory prediction (5s) | FDE: 5.833 -> 5.188 | Improved |
| Add global HDMap lane features to SocialGAN (HD-GAN) | SocialGAN · H-V2X trajectory prediction (1s) | ADE: 0.432 -> 0.427 | No clear effect |
| Add trajectory noise data augmentation (ablation) | SocialGAN / HD-GAN · H-V2X trajectory prediction with jittery observations | qualitative prediction error | Improved |
Closest papers by meaning
- V2X-Seq: A Large-Scale Sequential Dataset for Vehicle-Infrastructure Cooperative Perception and Forecasting (CVPR 2023 · similarity 0.90)
- UrbanIng-V2X: A Large-Scale Multi-Vehicle, Multi-Infrastructure Dataset Across Multiple Intersections for Cooperative Perception (NeurIPS 2025 · similarity 0.89)
- HoloVIC: Large-scale Dataset and Benchmark for Multi-Sensor Holographic Intersection and Vehicle-Infrastructure Cooperative (CVPR 2024 · similarity 0.89)
- V2X-Radar: A Multi-modal Dataset with 4D Radar for Cooperative Perception (NeurIPS 2025 · similarity 0.88)
- Map It Anywhere: Empowering BEV Map Prediction using Large-scale Public Datasets (NeurIPS 2024 · similarity 0.87)
- BEVSpread: Spread Voxel Pooling for Bird's-Eye-View Representation in Vision-based Roadside 3D Object Detection (CVPR 2024 · similarity 0.86)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
