English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Tools annual

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

In this paper, we explore methods of complicating self-supervised tasks forrepresentation learning. That is, we do severe damage to data and encourage anetwork to recover them. First, we complicate each of three powerfulself-supervised task candidates: jigsaw puzzle, inpainting, and colorization.In addition, we introduce a novel complicated self-supervised task called"Completing damaged jigsaw puzzles" which is puzzles with one piece missing andthe other pieces without color. We train a convolutional neural network notonly to solve the puzzles, but also generate the missing content and colorizethe puzzles. The recovery of the aforementioned damage pushes the network toobtain robust and general-purpose representations. We demonstrate thatcomplicating the self-supervised tasks improves their original versions andthat our final task learns more robust and transferable representationscompared to the previous methods, as well as the simple combination of ourcandidate tasks. Our approach achieves state-of-the-art performance in transferlearning on PASCAL classification and semantic segmentation.

Learning Image Representations by Completing Damaged Jigsaw Puzzles