Research map / Detection and segmentation / CNN features
Cross-View Transformers for Real-Time Map-View Semantic Segmentation
Check your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 cross-view attention with camera-aware positional embeddings | nuScenes Setting 2 vehicle map-view segmentation · nuScenes Setting 2 | IoU: 32.1 -> 36.0 | Helped |
| use cross-view attention with camera-aware positional embeddings | Argoverse vehicle map-view segmentation · Argoverse | IoU: 32.1 -> 35.2 | Helped |
| use cross-view attention with camera-aware positional embeddings | Argoverse driveable-area map-view segmentation · Argoverse | IoU: 65.4 -> 73.6 | Helped |
| use cross-view attention with camera-aware positional embeddings | nuScenes driveable-area map-view segmentation · nuScenes Setting 2 | IoU: 72.9 -> 74.3 | Helped |
| use cross-view attention with camera-aware positional embeddings | nuScenes Setting 1 vehicle map-view segmentation · nuScenes Setting 1 | IoU: 37.7 vs 37.5 | Mixed |
| add camera-aware positional embedding (ablation) | Cross-view Transformer · nuScenes Setting 2 | IoU: 31.0 -> 36.0 | Helped |
| add image features to attention keys (ablation) | Cross-view Transformer · nuScenes Setting 2 | IoU: 33.2 -> 36.0 | Helped |
| refine the map-view positional embedding (ablation) | Cross-view Transformer · nuScenes Setting 2 | IoU: 33.6 -> 36.0 | Helped |
| replace no positional embedding with camera-aware positional embedding (ablation) | Cross-view Transformer · nuScenes Setting 2 | IoU: 31.0 -> 36.0 | Helped |
| replace learned per-camera embeddings with camera-aware positional embeddings (ablation) | Cross-view Transformer · nuScenes Setting 2 | IoU: 34.4 -> 36.0 | Helped |
| use camera-aware positional embedding with random Fourier projection instead of linear projection (ablation) | Cross-view Transformer · nuScenes Setting 2 | IoU: 35.8 vs 36.0 | No clear effect |
Closest papers by meaning
- Rethinking Decoders for Transformer-based Semantic Segmentation: A Compression Perspective (NeurIPS 2024 · similarity 0.87)
- Understanding Multi-View Transformers (NeurIPS 2026 · similarity 0.87)
- EAGLE: Efficient Adaptive Geometry-based Learning in Cross-view Understanding (NeurIPS 2024 · similarity 0.87)
- Win-Win: Training High-Resolution Vision Transformers from Two Windows (ICLR 2024 · similarity 0.87)
- UniDSeg: Unified Cross-Domain 3D Semantic Segmentation via Visual Foundation Models Prior (NeurIPS 2024 · similarity 0.87)
- Large Spatial Model: End-to-end Unposed Images to Semantic 3D (NeurIPS 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.