AI Paper Atlas

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

SIFT Meets CNN: A Decade Survey of Instance Retrieval

Liang Zheng, Yi Yang, Qi Tian

TPAMI 2017 · Label-efficient and robust learning › Recognition and long-tail · approach: Correspondence and causal representation learning · 776 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
rotate Holidays images to uprightretrieval methods · HolidaysmAP: 2-3% mAP improvementImproved
replace pre-trained VGG16 with fine-tuned ResNet101 in Deep RetrievalDeep Retrieval · Oxford5kmAP: 83.1 -> 86.1Improved
add Hamming embedding with WGC to a large-codebook AKM baselineBoW retrieval · Oxford5kmAP: 49.3 -> 61.5Improved
add burstiness handling to HE+WGCHE+WGC · Oxford5kmAP: 61.5 -> 64.7Improved
add spatial verification/query expansion to Fine VocabularyFine Voc. · Oxford5kmAP: 74.2 -> 84.9Improved
fine-tune CNN on landmark data with MAC layer (SiaMAC) instead of pre-trained R-MACVGG16 pooled features · Oxford5kmAP: 66.9 -> 77.0Improved
increase HE signature length from 64 to 128 bitsHamming embedding · Oxford5kaccuracyImproved
increase HE signature to 256 bitsHamming embedding · Oxford5krecallHurt
increase codebook size in large/medium SIFT codebooksSIFT BoW · Oxford5kmAPMixed
add HE to medium-sized codebooksSIFT BoW · Oxford5kmAP stabilityImproved
reduce compact vector dimension below 256 or 128compact representations · Oxford5k/HolidaysmAP: 256, 128Hurt
fine-tune CNN on Landmarks datasetAlexNet retrieval · UkbenchmAPHurt

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