AI Paper Atlas

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

SVDNet for Pedestrian Retrieval

Yifan Sun, Liang Zheng, Weijian Deng, Shengjin Wang

ICCV 2017 · Label-efficient and robust learning › Recognition and long-tail · approach: Supervised person re-identification · 836 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
add SVDNet restraint-relaxation trainingCaffeNet FC6 · Market-1501rank-1 (%): 55.3 -> 80.5Improved
add SVDNet restraint-relaxation trainingCaffeNet FC6 · Market-1501mAP (%): 30.4 -> 55.9Improved
add SVDNet restraint-relaxation trainingResNet-50 Pool5 · Market-1501rank-1 (%): 73.8 -> 82.3Improved
add SVDNet restraint-relaxation trainingCaffeNet FC6 · CUHK03rank-1 (%): 38.6 -> 68.5Improved
add SVDNet restraint-relaxation trainingCaffeNet FC6 · DukeMTMC-reIDrank-1 (%): 46.9 -> 67.6Improved
add SVDNet restraint-relaxation trainingResNet-50 Pool5 · CUHK03rank-1 (%): 66.2 -> 81.8Improved
add SVDNet restraint-relaxation trainingResNet-50 Pool5 · DukeMTMC-reIDrank-1 (%): 65.5 -> 76.7Improved
replace W with US (SVD) at Step 1, no further trainingCaffeNet FC7 Step-0 model · Market-1501rank-1 (%): 63.6 -> 63.6No clear effect
replace W with U instead of US at Step 1 (ablation)CaffeNet FC7 Step-0 model · Market-1501rank-1 (%): 63.6 -> 61.7Hurt
replace W with QD instead of US at Step 1 (ablation)CaffeNet FC7 Step-0 model · Market-1501rank-1 (%): 63.6 -> 61.6Hurt
use US (keep singular values) instead of U in RRI (ablation)SVDNet CaffeNet FC6 · Market-1501rank-1 (%): 76.2 -> 80.5Improved
use US instead of UV^T in RRI (ablation)SVDNet CaffeNet FC6 · Market-1501rank-1 (%): 57.4 -> 80.5Improved

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