Beyond Finite Data: Towards Data-free Out-of-distribution Generalization via Extrapolation

• URL: http://arxiv.org/abs/2403.05523v2
• Date: Mon, 11 Mar 2024 04:58:52 GMT
• Title: Beyond Finite Data: Towards Data-free Out-of-distribution Generalization via Extrapolation
• Authors: Yijiang Li, Sucheng Ren, Weipeng Deng, Yuzhi Xu, Ying Gao, Edith Ngai and Haohan Wang
• Abstract summary: Humans can easily extrapolate novel domains, thus, an intriguing question arises: How can neural networks extrapolate like humans and achieve OOD generalization? We introduce a novel approach to domain extrapolation that leverages reasoning ability and the extensive knowledge encapsulated within large language models (LLMs) to synthesize entirely new domains. Our methods exhibit commendable performance in this setting, even surpassing the supervised setting by approximately 1-2% on datasets such as VLCS.
• Score: 19.944946262284123
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Out-of-distribution (OOD) generalization is a favorable yet challenging property for deep neural networks. The core challenges lie in the limited availability of source domains that help models learn an invariant representation from the spurious features. Various domain augmentation have been proposed but largely rely on interpolating existing domains and frequently face difficulties in creating truly "novel" domains. Humans, on the other hand, can easily extrapolate novel domains, thus, an intriguing question arises: How can neural networks extrapolate like humans and achieve OOD generalization? We introduce a novel approach to domain extrapolation that leverages reasoning ability and the extensive knowledge encapsulated within large language models (LLMs) to synthesize entirely new domains. Starting with the class of interest, we query the LLMs to extract relevant knowledge for these novel domains. We then bridge the gap between the text-centric knowledge derived from LLMs and the pixel input space of the model using text-to-image generation techniques. By augmenting the training set of domain generalization datasets with high-fidelity, photo-realistic images of these new domains, we achieve significant improvements over all existing methods, as demonstrated in both single and multi-domain generalization across various benchmarks. With the ability to extrapolate any domains for any class, our method has the potential to learn a generalized model for any task without any data. To illustrate, we put forth a much more difficult setting termed, data-free domain generalization, that aims to learn a generalized model in the absence of any collected data. Our empirical findings support the above argument and our methods exhibit commendable performance in this setting, even surpassing the supervised setting by approximately 1-2\% on datasets such as VLCS.

Related papers

• Domain Expansion and Boundary Growth for Open-Set Single-Source Domain Generalization [70.02187124865627]
  Open-set single-source domain generalization aims to use a single-source domain to learn a robust model that can be generalized to unknown target domains. We propose a novel learning approach based on domain expansion and boundary growth to expand the scarce source samples. Our approach can achieve significant improvements and reach state-of-the-art performance on several cross-domain image classification datasets.
  arXiv  Detail & Related papers  (2024-11-05T09:08:46Z)
• Learning to Generalize Unseen Domains via Multi-Source Meta Learning for Text Classification [71.08024880298613]
  We study the multi-source Domain Generalization of text classification. We propose a framework to use multiple seen domains to train a model that can achieve high accuracy in an unseen domain.
  arXiv  Detail & Related papers  (2024-09-20T07:46:21Z)
• Overcoming Data Inequality across Domains with Semi-Supervised Domain Generalization [4.921899151930171]
  We propose a novel algorithm, ProUD, which can effectively learn domain-invariant features via domain-aware prototypes. Our experiments on three different benchmark datasets demonstrate the effectiveness of ProUD.
  arXiv  Detail & Related papers  (2024-03-08T10:49:37Z)
• Normalization Perturbation: A Simple Domain Generalization Method for Real-World Domain Shifts [133.99270341855728]
  Real-world domain styles can vary substantially due to environment changes and sensor noises. Deep models only know the training domain style. We propose Normalization Perturbation to overcome this domain style overfitting problem.
  arXiv  Detail & Related papers  (2022-11-08T17:36:49Z)
• INDIGO: Intrinsic Multimodality for Domain Generalization [26.344372409315177]
  We study how multimodal information can be leveraged in an "intrinsic" way to make systems generalize under unseen domains. We propose IntriNsic multimodality for DomaIn GeneralizatiOn (INDIGO)
  arXiv  Detail & Related papers  (2022-06-13T05:41:09Z)
• Federated and Generalized Person Re-identification through Domain and Feature Hallucinating [88.77196261300699]
  We study the problem of federated domain generalization (FedDG) for person re-identification (re-ID) We propose a novel method, called "Domain and Feature Hallucinating (DFH)", to produce diverse features for learning generalized local and global models. Our method achieves the state-of-the-art performance for FedDG on four large-scale re-ID benchmarks.
  arXiv  Detail & Related papers  (2022-03-05T09:15:13Z)
• Towards Data-Free Domain Generalization [12.269045654957765]
  How can knowledge contained in models trained on different source data domains be merged into a single model that generalizes well to unseen target domains? Prior domain generalization methods typically rely on using source domain data, making them unsuitable for private decentralized data. We propose DEKAN, an approach that extracts and fuses domain-specific knowledge from the available teacher models into a student model robust to domain shift.
  arXiv  Detail & Related papers  (2021-10-09T11:44:05Z)
• COLUMBUS: Automated Discovery of New Multi-Level Features for Domain Generalization via Knowledge Corruption [12.555885317622131]
  We address the challenging domain generalization problem, where a model trained on a set of source domains is expected to generalize well in unseen domains without exposure to their data. We propose Columbus, a method that enforces new feature discovery via a targeted corruption of the most relevant input and multi-level representations of the data.
  arXiv  Detail & Related papers  (2021-09-09T14:52:05Z)
• Inferring Latent Domains for Unsupervised Deep Domain Adaptation [54.963823285456925]
  Unsupervised Domain Adaptation (UDA) refers to the problem of learning a model in a target domain where labeled data are not available. This paper introduces a novel deep architecture which addresses the problem of UDA by automatically discovering latent domains in visual datasets. We evaluate our approach on publicly available benchmarks, showing that it outperforms state-of-the-art domain adaptation methods.
  arXiv  Detail & Related papers  (2021-03-25T14:33:33Z)
• Learning to Generate Novel Domains for Domain Generalization [115.21519842245752]
  This paper focuses on the task of learning from multiple source domains a model that generalizes well to unseen domains. We employ a data generator to synthesize data from pseudo-novel domains to augment the source domains. Our method, L2A-OT, outperforms current state-of-the-art DG methods on four benchmark datasets.
  arXiv  Detail & Related papers  (2020-07-07T09:34:17Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.