Research map / Architectures and efficiency / CNN design
Efficient Neighbourhood Consensus Networks via Submanifold Sparse Convolutions
Ignacio Rocco, Relja Arandjelović, Josef Sivic
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 Sparse-NCNet (sparse 4D correlation with submanifold sparse convolutions) instead of dense NCNet | NCNet · HPatches Sequences | time, GPU memory, MMA: up to 10x faster and up to 20x less memory | Improved |
| use Sparse-NCNet instead of NCNet, both with hard relocalisation | Sparse-NCNet · InLoc | speed and memory: almost 7x faster and 6.5x less memory | Improved |
| add soft relocalisation after hard relocalisation (H+S vs H) (ablation) | Sparse-NCNet · HPatches Sequences | MMA | Improved |
| increase feature resolution to 200x150 (enabled by sparsity) (ablation) | Sparse-NCNet · HPatches Sequences | MMA, speed, memory: 30% faster and 3x more memory efficient than best NCNet variant | Improved |
| use two-stage relocalisation (H+S) at 200x150 resolution (ablation) | Sparse-NCNet · InLoc | localisation accuracy | Improved |
| use Sparse-NCNet instead of D2-Net / R2D2 | Sparse-NCNet · Aachen Day-Night | correctly localised queries: 68.4 vs 66.3 (1.0m, 5deg); 86.7 vs 88.8 (5.0m, 10deg) | Mixed |
Closest papers by meaning
- Neighbourhood Consensus Networks (NeurIPS 2018 · similarity 0.91)
- NCNet: Neighbourhood Consensus Networks for Estimating Image Correspondences (TPAMI 2020 · similarity 0.90)
- Correspondence Networks With Adaptive Neighbourhood Consensus (CVPR 2020 · similarity 0.90)
- DeMo: Deep Motion Field Consensus with Learnable Kernels for Two-view Correspondence Learning (AAAI 2025 · similarity 0.84)
- BCLNet: Bilateral Consensus Learning for Two-View Correspondence Pruning (AAAI 2024 · similarity 0.84)
- Role of Locality and Weight Sharing in Image-Based Tasks: A Sample Complexity Separation between CNNs, LCNs, and FCNs (ICLR 2024 · similarity 0.84)
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.
