ML p(r)ior | Unsupervised High-level Feature Learning by Ensemble Projection for Semi-supervised Image Classification and Image Clustering

Unsupervised High-level Feature Learning by Ensemble Projection for Semi-supervised Image Classification and Image Clustering

2016-02-02
1602.00955 | cs.CV
This paper investigates the problem of image classification with limited or no annotations, but abundant unlabeled data. The setting exists in many tasks such as semi-supervised image classification, image clustering, and image retrieval. Unlike previous methods, which develop or learn sophisticated regularizers for classifiers, our method learns a new image representation by exploiting the distribution patterns of all available data for the task at hand. Particularly, a rich set of visual prototypes are sampled from all available data, and are taken as surrogate classes to train discriminative classifiers; images are projected via the classifiers; the projected values, similarities to the prototypes, are stacked to build the new feature vector. The training set is noisy. Hence, in the spirit of ensemble learning we create a set of such training sets which are all diverse, leading to diverse classifiers. The method is dubbed Ensemble Projection (EP). EP captures not only the characteristics of individual images, but also the relationships among images. It is conceptually simple and computationally efficient, yet effective and flexible. Experiments on eight standard datasets show that: (1) EP outperforms previous methods for semi-supervised image classification; (2) EP produces promising results for self-taught image classification, where unlabeled samples are a random collection of images rather than being from the same distribution as the labeled ones; and (3) EP improves over the original features for image clustering. The code of the method is available on the project page.
PDF

Highlights - Most important sentences from the article

Login to like/save this paper, take notes and configure your recommendations

Related Articles

2018-05-05

Neural net classifiers trained on data with annotated class labels can also capture apparent visual … show more
PDF

Highlights - Most important sentences from the article

2017-11-03

Image descriptors based on activations of Convolutional Neural Networks (CNNs) have become dominant … show more
PDF

Highlights - Most important sentences from the article

2016-04-08

Convolutional Neural Networks (CNNs) achieve state-of-the-art performance in many computer vision ta… show more
PDF

Highlights - Most important sentences from the article

2014-03-23

Recent results indicate that the generic descriptors extracted from the convolutional neural network… show more
PDF

Highlights - Most important sentences from the article

2018-01-16
1801.05365 | cs.CV

We propose a deep learning-based solution for the problem of feature learning in one-class classific… show more
PDF

Highlights - Most important sentences from the article

2017-09-11

Low-shot learning methods for image classification support learning from sparse data. We extend thes… show more
PDF

Highlights - Most important sentences from the article

2017-01-26
1701.07717 | cs.CV

The main contribution of this paper is a simple semi-supervised pipeline that only uses the original… show more
PDF

Highlights - Most important sentences from the article

2017-12-04

Suffering from the extreme training data imbalance between seen and unseen classes, most of existing… show more
PDF

Highlights - Most important sentences from the article

2017-05-30

Remote sensing image classification is a fundamental task in remote sensing image processing. Remote… show more
PDF

Highlights - Most important sentences from the article

2018-12-19

Supervised deep learning algorithms have enabled significant performance gains in medical image clas… show more
PDF

Highlights - Most important sentences from the article

2015-06-16

In histopathological image analysis, feature extraction for classification is a challenging task due… show more
PDF

Highlights - Most important sentences from the article

2017-05-01

This paper presents a new supervised classification algorithm for remotely sensed hyperspectral imag… show more
PDF

Highlights - Most important sentences from the article

2016-09-09

Image classification is one of the main research problems in computer vision and machine learning. S… show more
PDF

Highlights - Most important sentences from the article

2019-04-17

In daily life, graphic symbols, such as traffic signs and brand logos, are ubiquitously utilized aro… show more
PDF

Highlights - Most important sentences from the article

2019-02-17

For many applications the collection of labeled data is expensive laborious. Exploitation of unlabel… show more
PDF

Highlights - Most important sentences from the article

2017-07-03

Due to the importance of zero-shot learning, i.e. classifying images where there is a lack of labele… show more