Machine learning with limited data
arXiv:2101.11461
Abstract
Thanks to the availability of powerful computing resources, big data and deep learning algorithms, we have made great progress on computer vision in the last few years. Computer vision systems begin to surpass humans in some tasks, such as object recognition, object detection, face recognition and pose estimation. Lots of computer vision algorithms have been deployed to real world applications and started to improve our life quality. However, big data and labels are not always available. Sometimes we only have very limited labeled data, such as medical images which requires experts to label them. In this paper, we study few shot image classification, in which we only have very few labeled data. Machine learning with little data is a big challenge. To tackle this challenge, we propose two methods and test their effectiveness thoroughly. One method is to augment image features by mixing the style of these images. The second method is applying spatial attention to explore the relations between patches of images. We also find that domain shift is a critical issue in few shot learning when the training domain and testing domain are different. So we propose a more realistic cross-domain few-shot learning with unlabeled data setting, in which some unlabeled data is available in the target domain. We propose two methods in this setting. Our first method transfers the style information of the unlabeled target dataset to the samples in the source dataset and trains a model with stylized images and original images. Our second method proposes a unified framework to fully utilize all the data. Both of our methods surpass the baseline method by a large margin.
References in corpus (27)
- An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale
- A Simple Framework for Contrastive Learning of Visual Representations
- mixup: Beyond Empirical Risk Minimization
- Language Models are Few-Shot Learners
- Unpaired Image-to-Image Translation using Cycle-Consistent Adversarial Networks
- Matching Networks for One Shot Learning
- Data Augmentation Generative Adversarial Networks
- Generalizing from a Few Examples: A Survey on Few-Shot Learning
- AutoAugment: Learning Augmentation Policies from Data
- Manifold Mixup: Better Representations by Interpolating Hidden States
- Adversarial Discriminative Domain Adaptation
- Self-paced Contrastive Learning with Hybrid Memory for Domain Adaptive Object Re-ID
- SimpleShot: Revisiting Nearest-Neighbor Classification for Few-Shot Learning
- A Closer Look at Few-shot Classification
- Unsupervised Learning via Meta-Learning
- Rethinking Few-Shot Image Classification: a Good Embedding Is All You Need?
- Dynamic Few-Shot Visual Learning without Forgetting
- Low-Shot Learning from Imaginary Data
- Unsupervised Meta-Learning For Few-Shot Image Classification
- Moment Matching for Multi-Source Domain Adaptation
- DeepEMD: Differentiable Earth Mover's Distance for Few-Shot Learning
- Image Deformation Meta-Networks for One-Shot Learning
- A Baseline for Few-Shot Image Classification
- Few-Shot Learning via Saliency-guided Hallucination of Samples
- A Broader Study of Cross-Domain Few-Shot Learning
- Low-Shot Learning with Imprinted Weights
- Disjoint Label Space Transfer Learning with Common Factorised Space