Abstract:Few-shot unsupervised domain adaptation (FS-UDA) utilizes few-shot labeled source domain data to realize effective classification in unlabeled target domain. However, current FS-UDA methods are still suffer from two issues: 1) the data from different domains can not be effectively aligned by few-shot labeled data due to the large domain gaps, 2) it is unstable and time-consuming to generalize to new FS-UDA <a class="link-external link-http" href="http://tasks.To" rel="external noopener nofollow">this http URL</a> address this issue, we put forward a novel Efficient Meta Prompt Learning Framework for FS-UDA. Within this framework, we use pre-trained CLIP model as the feature learning base model. First, we design domain-shared prompt learning vectors composed of virtual tokens, which mainly learns the meta knowledge from a large number of meta tasks to mitigate domain gaps. Secondly, we also design a task-shared prompt learning network to adaptively learn specific prompt vectors for each task, which aims to realize fast adaptation and task generalization. Thirdly, we learn a task-specific cross-domain alignment projection and a task-specific classifier with closed-form solutions for each meta task, which can efficiently adapt the model to new tasks in one step. The whole learning process is formulated as a bilevel optimization problem, and a good initialization of model parameters is learned through meta-learning. Extensive experimental study demonstrates the promising performance of our framework on benchmark datasets. Our method has the large improvement of at least 15.4% on 5-way 1-shot and 8.7% on 5-way 5-shot, compared with the state-of-the-art methods. Also, the performance of our method on all the test tasks is more stable than the other methods.

Convert Cross-Domain Classification into Few-Shot Learning: A Unified Prompt-Tuning Framework for Unsupervised Domain Adaptation

Adaptive Parametric Prototype Learning for Cross-Domain Few-Shot Classification

Learning a Target-Dependent Classifier for Cross-Domain Semantic Segmentation: Fine-tuning Versus Meta-Learning

Cross-domain Few-Shot Classification Through Feature Confusion

Cross-Domain Few-Shot Classification via Dense-Sparse-Dense Regularization

Domain Adaptation via Prompt Learning

Cross-domain error minimization for unsupervised domain adaptation

CrossTune: Black-Box Few-Shot Classification with Label Enhancement

Discriminative Sample-Guided and Parameter-Efficient Feature Space Adaptation for Cross-Domain Few-Shot Learning

Cross-Domain Few-Shot Classification Via Adversarial Task Augmentation

Prototypical Cross-domain Self-supervised Learning for Few-shot Unsupervised Domain Adaptation

HybridPrompt: Domain-Aware Prompting for Cross-Domain Few-Shot Learning

EMPL: A novel Efficient Meta Prompt Learning Framework for Few-shot Unsupervised Domain Adaptation

Exploring Cross-Domain Few-Shot Classification via Frequency-Aware Prompting

Cross-Domain Few-Shot Classification via Learned Feature-Wise Transformation

Cross-Level Distillation and Feature Denoising for Cross-Domain Few-Shot Classification

Bayesian Sharpness-Aware Prompt Tuning for Cross-Domain Few-shot Learning.

Cross-domain Few-shot Segmentation with Transductive Fine-tuning

Cross-Domain Few-Shot Learning via Adaptive Transformer Networks

Cross-domain Self-supervised Learning for Domain Adaptation with Few Source Labels