Unsupervised Meta-Learning For Few-Shot Image Classification

Khodadadeh, Siavash; Bölöni, Ladislau; Shah, Mubarak

doi:10.48550/arxiv.1811.11819

Cited by 10 publications

(17 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…20-way Acc. 1-shot 5-shot 1-shot 5-shot UMTRA (Khodadadeh et al, 2018) 77 The accuracy with std of our model is :33.77% ± 0.70%, 45.03% ± 0.73%, 53.35% ± 0.59%, 56.72% ± 0.67% on 5-way 1-shot, 5-way 5-shot, 5-way 20-shot, 5-way 50-shot, respectively.…”

Section: Results On Omniglotmentioning

confidence: 74%

“…Table 1 presents the performances of our model on the Omniglot dataset compared with other methods. We note that using the triplet loss, our model already outperforms other state-of-the-art unsupervised few-shot learning methods, including CACTUs (Hsu et al, 2018), UMTRA (Khodadadeh et al, 2018), and AAL (Antoniou & Storkey, 2019), to a large extend. Using the prototype loss, the performance of our model is further improved.…”

Section: Results On Omniglotmentioning

confidence: 77%

“…This is different from our model, in term of that we carry out progressive clustering and episodic training concurrently, and the two processes facilitate with each other. Khodadadeh et al (2018) proposed a method called UMTRA, which utilizes the statistical diversity properties and domain-specific augmentations to generate training and validation data. Antoniou & Storkey (2019) proposed a similar model called AAL, which uses data augmentations of the unlabeled support set to generate the query data.…”

Section: Related Workmentioning

confidence: 99%

See 2 more Smart Citations

Unsupervised Few-shot Learning via Self-supervised Training

Ji,

Zou,

Huang

et al. 2019

Preprint

View full text Add to dashboard Cite

Learning from limited exemplars (few-shot learning) is a fundamental, unsolved problem that has been laboriously explored in the machine learning community. However, current few-shot learners are mostly supervised and rely heavily on a large amount of labeled examples. Unsupervised learning is a more natural procedure for cognitive mammals and has produced promising results in many machine learning tasks. In the current study, we develop a method to learn an unsupervised few-shot learner via self-supervised training (UFLST), which can effectively generalize to novel but related classes. The proposed model consists of two alternate processes, progressive clustering and episodic training. The former generates pseudo-labeled training examples for constructing episodic tasks; and the later trains the few-shot learner using the generated episodic tasks which further optimizes the feature representations of data. The two processes facilitate with each other, and eventually produce a high quality few-shot learner. Using the benchmark dataset Omniglot and Mini-ImageNet, we show that our model outperforms other unsupervised few-shot learning methods. Using the benchmark dataset Mar-ket1501, we further demonstrate the feasibility of our model to a real-world application on person re-identification.

show abstract

Section: Results On Omniglotmentioning

confidence: 74%

Section: Results On Omniglotmentioning

confidence: 77%

Section: Related Workmentioning

confidence: 99%

See 1 more Smart Citation

Unsupervised Few-shot Learning via Self-supervised Training

Ji,

Zou,

Huang

et al. 2019

Preprint

View full text Add to dashboard Cite

show abstract

“…With more videos flourishing on the internet, recognizing human actions from videos [37,50,9,21,41,51,52] has drawn increasing attention in computer vision community. Recently, thanks to the strong capability of simultaneously capturing both spatial and temporal representations, 3DC-NNs have been widely and successfully explored in many video understanding tasks [4,5,14,42].…”

Section: Introductionmentioning

confidence: 99%

Self-Supervised Spatiotemporal Feature Learning via Video Rotation Prediction

Jing,

Yang,

Liu

et al. 2018

Preprint

116

View full text Add to dashboard Cite

The success of deep neural networks generally requires a vast amount of training data to be labeled, which is expensive and unfeasible in scale, especially for video collections. To alleviate this problem, in this paper, we propose 3DRotNet: a fully self-supervised approach to learn spatiotemporal features from unlabeled videos. A set of rotations are applied to all videos, and a pretext task is defined as prediction of these rotations. When accomplishing this task, 3DRotNet is actually trained to understand the semantic concepts and motions in videos. In other words, it learns a spatiotemporal video representation, which can be transferred to improve video understanding tasks in small datasets. Our extensive experiments successfully demonstrate the effectiveness of the proposed framework on action recognition, leading to significant improvements over the state-of-the-art self-supervised methods. With the selfsupervised pre-trained 3DRotNet from large datasets, the recognition accuracy is boosted up by 20.4% on UCF101 and 16.7% on HMDB51 respectively, compared to the models trained from scratch.

show abstract

“…Recently some researchers start to study this problem. UMTRA (Khodadadeh, Bölöni, and Shah, 2018) is an unsupervised metalearning method with tasks constructed by random sampling and augmentation.…”

Section: Unsupervised Meta Learningmentioning

confidence: 99%

Machine learning with limited data

Yao

2021

Preprint

View full text Add to dashboard Cite

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 crossdomain 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.

show abstract

Unsupervised Meta-Learning For Few-Shot Image Classification

Cited by 10 publications

References 0 publications

Unsupervised Few-shot Learning via Self-supervised Training

Unsupervised Few-shot Learning via Self-supervised Training

Self-Supervised Spatiotemporal Feature Learning via Video Rotation Prediction

Machine learning with limited data

Contact Info

Product

Resources

About