AI Paper Atlas

Research map / Label-efficient and robust learning / Zero-/few-shot and meta-learning

TransVLAD: Focusing on Locally Aggregated Descriptors for Few-Shot Learning

Haoquan Li, Laoming Zhang, Daoan Zhang, Lang Fu, Peng Yang, Jianguo Zhang

ECCV 2022 · Label-efficient and robust learning › Zero-/few-shot and meta-learning · approach: Embedding adaptation for few-shot classification

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 all four components (MAE pre-training, NeXtVLAD, masked fine-tuning, soft focal loss) (ablation)ViT-B trained directly · miniImageNet 5-way 1-shotaccuracy (%): 34.55 -> 68.24Improved
add all four components (ablation)ViT-B trained directly · miniImageNet 5-way 5-shotaccuracy (%): 45.73 -> 84.42Improved
add masking operation in fine-tuning (ablation)TransVLAD · miniImageNet 5-way 1-shotaccuracy (%): +2.8%Improved
mask 70% of patches in fine-tuning (ablation)TransVLAD · training speedaccuracy (%): more than 3 times fasterImproved
add NeXtVLAD instead of mean pooling (ablation)TransVLAD · miniImageNetaccuracy (%)Improved
add MAE pre-training before fine-tuning (ablation)ViT transformer · miniImageNetaccuracy (%)Improved
replace cross-entropy with soft focal loss (ablation)TransVLAD · miniImageNetaccuracy (%)Improved
set focusing parameter gamma to 2 (ablation)TransVLAD · miniImageNet validationaccuracy (%): best at gamma = 2Improved
set masking ratio to 70% (ablation)TransVLAD · miniImageNet validationaccuracy (%): best at 70%Improved
train on larger tieredImageNetTransVLAD vs SSFormers · tieredImageNet 5-way 1-shotaccuracy (%): 72.52 -> 77.20Improved
transfer from miniImageNet to CUBTransVLAD vs S2M2 · CUB cross-domain 5-way 1-shotaccuracy (%): 48.42 -> 50.45Improved
train on CUB and test on other datasetsTransVLAD · CUB->miniImageNetaccuracy (%): 68.24 -> 39.36Hurt

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