Research map / Detection and segmentation / 3D detection and driving
Pixel-Aligned Recurrent Queries for Multi-View 3D Object Detection
Yiming Xie, Huaizu Jiang, Georgia Gkioxari, Julian Straub
This paper tried a common change: Add positional encoding
add ray positional encodings to image features · ScanNet · F1 @IoU>0.5: 28.2 -> 30.0 Improved
What 646 accepted papers report for this change
- 525 improved
- 58 no clear effect
- 7 mixed
- 56 hurt
- Learning-Augmented Scalable Linear Assignment Problem Optimization via Neural Dual Warm-Starts ICML 2026 · add positional encodings to features (d=13 -> d=21) · Block-structured, N=4096 · speedupImproved
- Mesh Based Simulations with Spatial and Temporal awareness ICML 2026 · use learned positional embedding, learnable relative bias or weighted adjacency instead of RoPE · Aneurysm · RMSEHurt
- Sparse ActionGen: Accelerating Diffusion Policy with Real-time Pruning ICML 2026 · add sinusoidal positional encoding to the pruner · Robomimic Tool Hang · success rate (%): 28 -> 54Improved
See this change in detailMatched by the wording of each paper's own result, read by AI from the full text. Papers report gains more often than failures, so this shows what gets published, not the chance a change works.
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 |
|---|---|---|---|
| add pixel-aligned appearance features to queries (ablation) | PARQ · ScanNet | F1 @IoU>0.5: 27.1 -> 30.0 | Improved |
| add recurrent refinement of 3D reference points (ablation) | PARQ · ScanNet | F1 @IoU>0.5: 27.6 -> 30.0 | Improved |
| include positional encoding of 3D reference point in query (ablation) | PARQ · ScanNet | F1 @IoU>0.5: 10.4 -> 30.0 | Improved |
| add ray positional encodings to image features (ablation) | PARQ · ScanNet | F1 @IoU>0.5: 28.2 -> 30.0 | Improved |
| use more views at inference than training (3) | PARQ · ScanNet | F1 @IoU>0.5: 30.0 -> 31.3 (7 views) | Improved |
| use more views at inference than training (3) | PETR and ImVoxelNet · ScanNet | F1 @IoU>0.5: PETR 25.6 -> 24.4 (7 views); ImVoxelNet 26.8 -> 23.3 (9 views) | No clear effect |
| reduce queries to 32 and iterations to 2 at inference | PARQ · ScanNet | inference time and F1: 290 ms -> 105 ms; F1 30.0 -> 28.2 at IoU>0.5 | Mixed |
| use PARQ rather than ImVoxelNet at loose IoU threshold | ScanNet IoU>0.25 · ScanNet | F1: 51.7 -> 51.1 | Hurt |
| use PARQ instead of PETR on ARKitScenes | multi-view 3D detection · ARKitScenes IoU>0.5 | F1: 17.5 -> 24.1 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 on most shared results.
- PETR — PETR: Position Embedding Transformation for Multi-View 3D Object Detection (ECCV 2022) · better on 7 of 7 shared results
- ODAM — ODAM: Object Detection, Association, and Mapping Using Posed RGB Video (ICCV 2021) · better on 1 of 1 shared result
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Point2Seq: Detecting 3D Objects As Sequences (CVPR 2022 · similarity 0.88)
- Towards Generalizable Multi-Camera 3D Object Detection via Perspective Rendering (AAAI 2025 · similarity 0.86)
- Far3D: Expanding the Horizon for Surround-View 3D Object Detection (AAAI 2024 · similarity 0.86)
- Tango3D: Towards Alignment for Global and Local 2D-3D Correspondence (NeurIPS 2026 · similarity 0.85)
- Multi-View Attentive Contextualization for Multi-View 3D Object Detection (CVPR 2024 · similarity 0.85)
- RefineAny3D: Depth Refinement as Semantic Alignment for Monocular 3D Detection (NeurIPS 2026 · similarity 0.85)
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.
