AI Paper Atlas

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

SelaVPR++: Towards Seamless Adaptation of Foundation Models for Efficient Place Recognition

Feng Lu, Tong Jin, Xiangyuan Lan, Lijun Zhang, Yunpeng Liu, Yaowei Wang, Chun Yuan

TPAMI 2025 · Label-efficient and robust learning › Recognition and long-tail · approach: Correspondence and causal representation learning · 15 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 MultiConv adapters in the memory-efficient parallel network instead of vanilla adapters (ablation)MemEfficient adaptation (DINOv2-B) · NordlandR@1: 75.0 -> 81.3Improved
use MultiConv adapters instead of token-channel alternating adapters (ablation)MemEfficient adaptation (224px) · NordlandR@1: 73.1 -> 75.2Improved
use memory-efficient MultiConv adaptation instead of full fine-tuning (ablation)DINOv2-B backbone · NordlandR@1: 73.7 -> 81.3Improved
use memory-efficient MultiConv adaptation instead of frozen backbone (ablation)DINOv2-B backbone · NordlandR@1: 53.2 -> 81.3Improved
use memory-efficient MultiConv adaptation instead of partial tuning (last 4 blocks) (ablation)DINOv2-B backbone · NordlandR@1: 77.4 -> 81.3Improved
use memory-efficient MultiConv adaptation instead of SelaVPR-global adapters (ablation)DINOv2-B backbone · NordlandR@1: 73.3 -> 81.3Improved
use memory-efficient MultiConv adaptation instead of full fine-tuning (ablation)DINOv2-L backbone · Tokyo24/7R@1: 88.9 -> 96.2Improved
use memory-efficient MultiConv adaptation instead of partial tuning (ablation)DINOv2-L backbone · NordlandR@1: 78.8 -> 81.4Improved
freezing backbone instead of MultiConv adaptation (ablation)DINOv2-L backbone · Tokyo24/7R@5: 98.4 -> 98.4No clear effect
replace SelaVPR-global adaptation with MemEfficient-MCA to cut training memory (ablation)DINOv2-L backbone · GPU memoryratio of memory: 14.6% of SelaVPR-global, 11.5% of Full-tuningImproved
use MultiConv adaptation instead of partial tuning (DINOv2-B) (ablation)DINOv2-B backbone · GSV-Cities trainingmin/epoch: 6.85 vs 6.95No clear effect
apply only 4 adapters instead of 16 for DINOv2-L (ablation)MemEfficient-MCA · GSV-Cities trainingmin/epoch: 11.97 -> 9.60Improved

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