Research map / Detection and segmentation / CNN features
MTLDesc: Looking Wider to Describe Better
Changwei Wang, Rongtao Xu, Yuyang Zhang, Shibiao Xu, Weiliang Meng, Bin Fan, Xiaopeng Zhang
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 |
|---|---|---|---|
| add feature pyramid keypoint detection (ablation) | SuperPoint our impl. · HPatches @3 | MMA%: 67.51 -> 69.88 | Improved |
| add AGCA with gated map rather than AGCA without gated map (ablation) | Baseline + Context Augmentation · HPatches @3 | MMA%: 71.33 -> 72.28 | Improved |
| add AGCA (ablation) | Baseline · HPatches @3 | MMA%: 67.51 -> 72.28 | Improved |
| add DLCA (ablation) | Baseline · HPatches @3 | MMA%: 67.51 -> 71.25 | Improved |
| combine AGCA and DLCA (ablation) | Baseline · HPatches @3 | MMA%: 67.51 -> 73.14 | Improved |
| apply consistent attention in descriptor optimization (ablation) | Baseline + Context Augmentation · HPatches @3 | MMA%: 73.14 -> 74.94 | Improved |
| use attention-weighted descriptors during matching as well as optimization (ablation) | Baseline + Context Augmentation + Consistent Attention Weighting · HPatches @3 | MMA%: 74.94 -> 78.66 | Improved |
| use attention-weighted descriptors during matching (ablation) | Baseline + Context Augmentation + Consistent Attention Weighting · HPatches @3 | M.S.%: 50.67 -> 47.16 | Hurt |
| add consistent attention weighting (ablation) | MTLDesc · HPatches @3 | MMA%: 73.14 -> 78.66 | Improved |
| use MTLDesc rather than DISK (8K) | local feature matching · HPatches @3 | MMA%: 77.58 -> 78.66 | Improved |
| use MTLDesc rather than DISK | local feature description · Aachen Day-Night v1.1 | localized queries (0.25m,2°): 72.8 -> 74.3 | Improved |
| use MTLDesc rather than R2D2 | local feature description · InLoc DUC1 | localized queries (0.5m): 57.1 -> 61.6 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- R2D2 — R2D2: Reliable and Repeatable Detector and Descriptor (NeurIPS 2019) · better on 3 of 4 shared results
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
- Attention Weighted Local Descriptors (TPAMI 2023 · similarity 0.89)
- ContextDesc: Local Descriptor Augmentation With Cross-Modality Context (CVPR 2019 · similarity 0.86)
- D2-Net: A Trainable CNN for Joint Description and Detection of Local Features (CVPR 2019 · similarity 0.85)
- From Coarse to Fine: Robust Hierarchical Localization at Large Scale (CVPR 2019 · similarity 0.84)
- ASLFeat: Learning Local Features of Accurate Shape and Localization (CVPR 2020 · similarity 0.84)
- Focus on Local: Finding Reliable Discriminative Regions for Visual Place Recognition (AAAI 2025 · similarity 0.83)
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.
