Recall@k Surrogate Loss with Large Batches and Similarity Mixup
The result's identifiers
Result code in IS VaVaI
<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F68407700%3A21230%2F22%3A00362948" target="_blank" >RIV/68407700:21230/22:00362948 - isvavai.cz</a>
Result on the web
<a href="https://doi.org/10.1109/CVPR52688.2022.00735" target="_blank" >https://doi.org/10.1109/CVPR52688.2022.00735</a>
DOI - Digital Object Identifier
<a href="http://dx.doi.org/10.1109/CVPR52688.2022.00735" target="_blank" >10.1109/CVPR52688.2022.00735</a>
Alternative languages
Result language
angličtina
Original language name
Recall@k Surrogate Loss with Large Batches and Similarity Mixup
Original language description
This work focuses on learning deep visual representation models for retrieval by exploring the interplay between a new loss function, the batch size, and a new regularization approach. Direct optimization, by gradient descent, of an evaluation metric, is not possible when it is non-differentiable, which is the case for recall in retrieval. A differentiable surrogate loss for the recall is proposed in this work. Using an implementation that sidesteps the hardware constraints of the GPU memory, the method trains with a very large batch size, which is essential for metrics computed on the entire retrieval database. It is assisted by an efficient mixup regularization approach that operates on pairwise scalar similarities and virtually increases the batch size further. The suggested method achieves state-of-the-art performance in several image retrieval benchmarks when used for deep metric learning. For instance-level recognition, the method outperforms similar approaches that train using an approximation of average precision.
Czech name
—
Czech description
—
Classification
Type
D - Article in proceedings
CEP classification
—
OECD FORD branch
10201 - Computer sciences, information science, bioinformathics (hardware development to be 2.2, social aspect to be 5.8)
Result continuities
Project
Result was created during the realization of more than one project. More information in the Projects tab.
Continuities
P - Projekt vyzkumu a vyvoje financovany z verejnych zdroju (s odkazem do CEP)
Others
Publication year
2022
Confidentiality
S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů
Data specific for result type
Article name in the collection
Proceeding 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)
ISBN
978-1-6654-6946-3
ISSN
1063-6919
e-ISSN
2575-7075
Number of pages
10
Pages from-to
7492-7501
Publisher name
IEEE
Place of publication
Piscataway
Event location
New Orleans, Louisiana
Event date
Jun 19, 2022
Type of event by nationality
WRD - Celosvětová akce
UT code for WoS article
000870759100033