Weakly-supervised segmentation and unsupervised modeling of natural images

By: Kolesnikov, Alexander
Material type: TextTextPublisher: IST Austria 2018Online resources: Click here to access online
Contents:
Abstract
Acknowledgments
1 Introduction and Background
2 Semantic Image Segmentation with Image-Level Labels
3 Semantic Image Segmentation with Micro-Annotation
4 Probabilistic Autoregressive Image Modelling
5 Automatic Image Colorization
6 Conclusion and Future Work
Summary: Modern computer vision systems heavily rely on statistical machine learning models, which typically require large amounts of labeled data to be learned reliably. Moreover, very recently computer vision research widely adopted techniques for representation learning, which further increase the demand for labeled data. However, for many important practical problems there is relatively small amount of labeled data available, so it is problematic to leverage full potential of the representation learning methods. One way to overcome this obstacle is to invest substantial resources into producing large labelled datasets. Unfortunately, this can be prohibitively expensive in practice. In this thesis we focus on the alternative way of tackling the aforementioned issue. We concentrate on methods, which make use of weakly-labeled or even unlabeled data. Specifically, the first half of the thesis is dedicated to the semantic image segmentation task. We develop a technique, which achieves competitive segmentation performance and only requires annotations in a form of global image-level labels instead of dense segmentation masks. Subsequently, we present a new methodology, which further improves segmentation performance by leveraging tiny additional feedback from a human annotator. By using our methods practitioners can greatly reduce the amount of data annotation effort, which is required to learn modern image segmentation models. In the second half of the thesis we focus on methods for learning from unlabeled visual data. We study a family of autoregressive models for modeling structure of natural images and discuss potential applications of these models. Moreover, we conduct in-depth study of one of these applications, where we develop the state-of-the-art model for the probabilistic image colorization task.
List(s) this item appears in: IST Austria Thesis
Tags from this library: No tags from this library for this title. Log in to add tags.
    Average rating: 0.0 (0 votes)
Item type Current location Call number Status Date due Barcode Item holds
Book Book Library
Available AT-ISTA#001522
Total holds: 0

Thesis

Abstract

Acknowledgments

1 Introduction and Background

2 Semantic Image Segmentation with Image-Level Labels

3 Semantic Image Segmentation with Micro-Annotation

4 Probabilistic Autoregressive Image Modelling

5 Automatic Image Colorization

6 Conclusion and Future Work

Modern computer vision systems heavily rely on statistical machine learning models, which typically require large amounts of labeled data to be learned reliably. Moreover, very recently computer vision research widely adopted techniques for representation learning, which further increase the demand for labeled data. However, for many important practical problems there is relatively small amount of labeled data available, so it is problematic to leverage full potential of the representation learning methods. One way to overcome this obstacle is to invest substantial resources into producing large labelled datasets. Unfortunately, this can be prohibitively expensive in practice. In this thesis we focus on the alternative way of tackling the aforementioned issue. We concentrate on methods, which make use of weakly-labeled or even unlabeled data. Specifically, the first half of the thesis is dedicated to the semantic image segmentation task. We develop a technique, which achieves competitive segmentation performance and only requires annotations in a form of global image-level labels instead of dense segmentation masks. Subsequently, we present a new methodology, which further improves segmentation performance by leveraging tiny additional feedback from a human annotator. By using our methods practitioners can greatly reduce the amount of data annotation effort, which is required to learn modern image segmentation models. In the second half of the thesis we focus on methods for learning from unlabeled visual data. We study a family of autoregressive models for modeling structure of natural images and discuss potential applications of these models. Moreover, we conduct in-depth study of one of these applications, where we develop the state-of-the-art model for the probabilistic image colorization task.

There are no comments for this item.

to post a comment.

Powered by Koha