---
res:
  bibo_abstract:
  - Deep neural networks (DNNs) are vulnerable to small adversarial perturbations,
    which are tiny changes to the input data that appear insignificant but cause the
    model to produce drastically different outputs. Many defense methods require modifying
    model architectures during evaluation or performing test-time data purification.
    This not only introduces additional complexity but is often architecture-dependent.
    We show, however, that robust feature learning during training can significantly
    enhance DNN robustness. We propose MOREL, a multi-objective approach that aligns
    natural and adversarial features using cosine similarity and multi-positive contrastive
    losses to encourage similar features for same-class inputs. Extensive experiments
    demonstrate that MOREL significantly improves robustness against both white-box
    and black-box attacks. Our code is available at https://github.com/salomonhotegni/MOREL.@eng
  bibo_authorlist:
  - foaf_Person:
      foaf_givenName: Sedjro Salomon
      foaf_name: Hotegni, Sedjro Salomon
      foaf_surname: Hotegni
      foaf_workInfoHomepage: http://www.librecat.org/personId=97995
  - foaf_Person:
      foaf_givenName: Sebastian
      foaf_name: Peitz, Sebastian
      foaf_surname: Peitz
      foaf_workInfoHomepage: http://www.librecat.org/personId=47427
    orcid: 0000-0002-3389-793X
  bibo_doi: 10.1007/978-3-032-04558-4_35
  dct_date: 2026^xs_gYear
  dct_isPartOf:
  - http://id.crossref.org/issn/978-3-032-04558-4
  dct_language: eng
  dct_publisher: Springer Nature Switzerland@
  dct_subject:
  - own
  - own-conference
  dct_title: Enhancing Adversarial Robustness Through Multi-objective Representation
    Learning@
...