Adaptive SVM+: Learning with Privileged Information for Domain Adaptation

Abstract

Incorporating additional knowledge in the learning process can be beneficial for several computer vision and machine learning tasks. Whether privileged information originates from a source domain that is adapted to a target domain, or as additional features available at training time only, using such privileged (i.e., auxiliary) information is of high importance as it improves the recognition performance and generalization. However, both primary and privileged information are rarely derived from the same distribution, which poses an additional challenge to the recognition task. To address these challenges, we present a novel learning paradigm that leverages privileged information in a domain adaptation setup to perform visual recognition tasks. The proposed framework, named Adaptive SVM+, combines the advantages of both the learning using privileged information (LUPI) paradigm and the domain adaptation framework, which are naturally embedded in the objective function of a regular SVM. We demonstrate the effectiveness of our approach on the publicly available Animals with Attributes and INTERACT datasets and report state-of-the-art results in both of them.

Integration of mechanistic immunological knowledge into a machine learning pipeline improves predictions

“…robustness [19][20][21][22][23][24][25][26][27][28][29][30][31] . In biological sciences, examples of such knowledge integration include inference of biological networks 24,32 and causal pathway modelling 33,34 .…”

Curriculum learning of visual attribute clusters for multi-task classification

“…Pentina et al [52] introduced a curriculum learning-based approach to process multiple tasks in a sequence and developed a method to find the best order in which the tasks need to be learned. They proposed a data-dependent solution by introducing an upper-bound of the average expected error and employing an Adaptive SVM [53,54]. Such a learning process has the advantage of exploiting prior knowledge to improve subsequent classification tasks but it cannot scale up to many tasks since each subsequent task has to be learned individually.…”

Distraction-Aware Feature Learning for Human Attribute Recognition via Coarse-to-Fine Attention Mechanism

“…Existing investigations of HAR can be classified into three domains, i.e., clothing domain, surveillance domain, and general domain. The techniques in the clothing domain have received extensive attentions (Al-Halah, Stiefelhagen, and Grauman 2017;Sarafianos, Vrigkas, and Kakadiaris 2017;Liu et al 2016;Chen et al 2015) due to their potentials in commercial applications. This type of methods generally require the input images of high resolutions with persons at a small number of pre-defined poses, and fine-grained clothing style recognition is still challenging.…”