File(s) under permanent embargo
Beyond dataset bias: multi-task unaligned shared knowledge transfer
conference contribution
posted on 2023-06-08, 16:45 authored by Tatiana Tommasi, Novi QuadriantoNovi Quadrianto, Barbara Caputo, Christoph H LampertMany visual datasets are traditionally used to analyze the performance of different learning techniques. The evaluation is usually done within each dataset, therefore it is questionable if such results are a reliable indicator of true generalization ability. We propose here an algorithm to exploit the existing data resources when learning on a new multiclass problem. Our main idea is to identify an image representation that decomposes orthogonally into two subspaces: a part specific to each dataset, and a part generic to, and therefore shared between, all the considered source sets. This allows us to use the generic representation as un-biased reference knowledge for a novel classification task. By casting the method in the multi-view setting, we also make it possible to use different features for different databases. We call the algorithm MUST, Multitask Unaligned Shared knowledge Transfer. Through extensive experiments on five public datasets, we show that MUST consistently improves the cross-datasets generalization performance.
History
Publication status
- Published
File Version
- Published version
Journal
Proceedings of Computer vision - ACCV 2012: 11th Asian Conference on Computer Vision; Daejeon, Korea; 5-9 November 2012ISSN
0302-9743Publisher
Springer VerlagExternal DOI
Issue
7724Page range
1-15Pages
821.0Event name
11th Asian conference on computer vision (ACCV)Event location
Daejeon, KoreaEvent type
conferenceEvent date
5-9 November, 2012Book title
Computer vision – ACCV 2012: 11th Asian conference on computer vision, Daejeon, Korea, November 5-9, 2012, revised selected papers, Part IPlace of publication
Berlin; New YorkISBN
9783642373305Series
Lecture notes in computer scienceDepartment affiliated with
- Informatics Publications
Full text available
- No
Peer reviewed?
- Yes