AI Paper Atlas

Research map / Label-efficient and robust learning / Recognition and long-tail

TopicFM: Robust and Interpretable Topic-Assisted Feature Matching

Khang Truong Giang, Soohwan Song, Sungho Jo

AAAI 2023 · Label-efficient and robust learning › Recognition and long-tail · approach: Correspondence and causal representation learning · 62 citations (OpenAlex)

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 changedWhereMeasuredEffect
Use 50 topics instead of 10 (ablation)TopicFM · MegaDepthrelative pose estimation AUC (5°): 48.6 -> 49.1Improved
Increase topic count from 50 to 100 (ablation)TopicFM · MegaDepthrelative pose estimation AUC (5°): 49.1 -> 49.0No clear effect
Increase covisible topics from 2 to 6 (ablation)TopicFM · MegaDepthrelative pose estimation AUC (5°): 49.3 -> 52.7Improved
Increase covisible topics from 6 to 10 (ablation)TopicFM · MegaDepthrelative pose estimation AUC (5°): 52.7 -> 54.1Improved
Increase covisible topics from 10 to 12 (ablation)TopicFM · MegaDepthrelative pose estimation AUC (5°): 54.1 -> 53.7Hurt
Use topic-assisted matching rather than LoFTRimage matching · HPatcheshomography estimation AUC (3px): 65.9 -> 67.3Improved
Use topic-assisted matching rather than MatchFormerimage matching · MegaDepthrelative pose estimation AUC (10°): 69.4 -> 70.1Improved
Use topic-assisted matching rather than LoFTRimage matching · ScanNetrelative pose estimation AUC (20°): 50.62 -> 50.91Improved
Use TopicFM rather than LoFTRvisual localization · InLocoverall: 69.8 -> 70.9Improved
Use TopicFM rather than LoFTRfeature matching · 1200 × 896 runtime benchmarkruntime (ms): 500 -> 323Improved

Later papers that compared with this method

2 later accepted papers report TopicFM as a baseline in their results tables; 2 of them report a better number on most of the results they share with it.

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

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.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy