SoRA: Singular Value Decomposed Low-Rank Adaptation for Domain Generalizable Representation Learning

• URL: http://arxiv.org/abs/2412.04077v1
• Date: Thu, 05 Dec 2024 11:17:57 GMT
• Title: SoRA: Singular Value Decomposed Low-Rank Adaptation for Domain Generalizable Representation Learning
• Authors: Seokju Yun, Seunghye Chae, Dongheon Lee, Youngmin Ro,
• Abstract summary: Domain generalization aims to adapt a model using one or multiple source domains to ensure robust performance in unseen target domains. Existing PEFT methods struggle to strike a balance between preserving generalizable components of the pre-trained model and learning task-specific features. We introduce Singular Value Decomposed Low-Rank Adaptation (SoRA), an approach that selectively tunes minor singular components while keeping the residual parts frozen.
• Score: 6.262268096839562
• License:
• Abstract: Domain generalization (DG) aims to adapt a model using one or multiple source domains to ensure robust performance in unseen target domains. Recently, Parameter-Efficient Fine-Tuning (PEFT) of foundation models has shown promising results in the context of DG problem. Nevertheless, existing PEFT methods still struggle to strike a balance between preserving generalizable components of the pre-trained model and learning task-specific features. To gain insights into the distribution of generalizable components, we begin by analyzing the pre-trained weights through the lens of singular value decomposition. Building on these insights, we introduce Singular Value Decomposed Low-Rank Adaptation (SoRA), an approach that selectively tunes minor singular components while keeping the residual parts frozen. SoRA effectively retains the generalization ability of the pre-trained model while efficiently acquiring task-specific skills. Furthermore, we freeze domain-generalizable blocks and employ an annealing weight decay strategy, thereby achieving an optimal balance in the delicate trade-off between generalizability and discriminability. SoRA attains state-of-the-art results on multiple benchmarks that span both domain generalized semantic segmentation to domain generalized object detection. In addition, our methods introduce no additional inference overhead or regularization loss, maintain compatibility with any backbone or head, and are designed to be versatile, allowing easy integration into a wide range of tasks.

Related papers

• Learn to Preserve and Diversify: Parameter-Efficient Group with Orthogonal Regularization for Domain Generalization [28.977757627384165]
  Domain Domain (DG) aims to avoid the performance degradation of the model when the distribution shift between the limited training data and unseen test data occurs. Recently, foundation models with enormous parameters have been pre-trained with huge datasets, demonstrating strong generalization ability. Our framework achieves SOTA performance on five DG benchmarks, while only requiring training a small number of parameters without adding additional testing cost.
  arXiv  Detail & Related papers  (2024-07-21T07:50:49Z)
• Causality-inspired Latent Feature Augmentation for Single Domain Generalization [13.735443005394773]
  Single domain generalization (Single-DG) intends to develop a generalizable model with only one single training domain to perform well on other unknown target domains. Under the domain-hungry configuration, how to expand the coverage of source domain and find intrinsic causal features across different distributions is the key to enhancing the models' generalization ability. We propose a novel causality-inspired latent feature augmentation method for Single-DG by learning the meta-knowledge of feature-level transformation based on causal learning and interventions.
  arXiv  Detail & Related papers  (2024-06-10T02:42:25Z)
• LoRA-Ensemble: Efficient Uncertainty Modelling for Self-attention Networks [52.46420522934253]
  We introduce LoRA-Ensemble, a parameter-efficient deep ensemble method for self-attention networks. By employing a single pre-trained self-attention network with weights shared across all members, we train member-specific low-rank matrices for the attention projections. Our method exhibits superior calibration compared to explicit ensembles and achieves similar or better accuracy across various prediction tasks and datasets.
  arXiv  Detail & Related papers  (2024-05-23T11:10:32Z)
• Rethinking Multi-domain Generalization with A General Learning Objective [17.155829981870045]
  Multi-domain generalization (mDG) is universally aimed to minimize discrepancy between training and testing distributions. Existing mDG literature lacks a general learning objective paradigm. We propose to leverage a $Y$-mapping to relax the constraint.
  arXiv  Detail & Related papers  (2024-02-29T05:00:30Z)
• Mitigate Domain Shift by Primary-Auxiliary Objectives Association for Generalizing Person ReID [39.98444065846305]
  ReID models struggle in learning domain-invariant representation solely through training on an instance classification objective. We introduce a method that guides model learning of the primary ReID instance classification objective by a concurrent auxiliary learning objective on weakly labeled pedestrian saliency detection. Our model can be extended with the recent test-time diagram to form the PAOA+, which performs on-the-fly optimization against the auxiliary objective.
  arXiv  Detail & Related papers  (2023-10-24T15:15:57Z)
• Adaptive Domain Generalization via Online Disagreement Minimization [17.215683606365445]
  Domain Generalization aims to safely transfer a model to unseen target domains. AdaODM adaptively modifies the source model at test time for different target domains. Results show AdaODM stably improves the generalization capacity on unseen domains.
  arXiv  Detail & Related papers  (2022-08-03T11:51:11Z)
• META: Mimicking Embedding via oThers' Aggregation for Generalizable Person Re-identification [68.39849081353704]
  Domain generalizable (DG) person re-identification (ReID) aims to test across unseen domains without access to the target domain data at training time. This paper presents a new approach called Mimicking Embedding via oThers' Aggregation (META) for DG ReID.
  arXiv  Detail & Related papers  (2021-12-16T08:06:50Z)
• Calibrated Feature Decomposition for Generalizable Person Re-Identification [82.64133819313186]
  Calibrated Feature Decomposition (CFD) module focuses on improving the generalization capacity for person re-identification. A calibrated-and-standardized Batch normalization (CSBN) is designed to learn calibrated person representation.
  arXiv  Detail & Related papers  (2021-11-27T17:12:43Z)
• Towards Principled Disentanglement for Domain Generalization [90.9891372499545]
  A fundamental challenge for machine learning models is generalizing to out-of-distribution (OOD) data. We first formalize the OOD generalization problem as constrained optimization, called Disentanglement-constrained Domain Generalization (DDG) Based on the transformation, we propose a primal-dual algorithm for joint representation disentanglement and domain generalization.
  arXiv  Detail & Related papers  (2021-11-27T07:36:32Z)
• Learning Invariant Representations and Risks for Semi-supervised Domain Adaptation [109.73983088432364]
  We propose the first method that aims to simultaneously learn invariant representations and risks under the setting of semi-supervised domain adaptation (Semi-DA) We introduce the LIRR algorithm for jointly textbfLearning textbfInvariant textbfRepresentations and textbfRisks.
  arXiv  Detail & Related papers  (2020-10-09T15:42:35Z)
• Feature Alignment and Restoration for Domain Generalization and Adaptation [93.39253443415392]
  Cross domain feature alignment has been widely explored to pull the feature distributions of different domains in order to learn domain-invariant representations. We propose a unified framework termed Feature Alignment and Restoration (FAR) to simultaneously ensure high generalization and discrimination power of the networks. Experiments on multiple classification benchmarks demonstrate the high performance and strong generalization of our FAR framework for both domain generalization and unsupervised domain adaptation.
  arXiv  Detail & Related papers  (2020-06-22T05:08:13Z)

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.