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 Plus annual

Presents the access of premium content as premium feature

Premium Content

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 Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

The study of object representations in computer vision has primarily focusedon developing representations that are useful for image classification, objectdetection, or semantic segmentation as downstream tasks. In this work we aim tolearn object representations that are useful for control and reinforcementlearning (RL). To this end, we introduce Transporter, a neural networkarchitecture for discovering concise geometric object representations in termsof keypoints or image-space coordinates. Our method learns from raw videoframes in a fully unsupervised manner, by transporting learnt image featuresbetween video frames using a keypoint bottleneck. The discovered keypointstrack objects and object parts across long time-horizons more accurately thanrecent similar methods. Furthermore, consistent long-term tracking enables twonotable results in control domains -- (1) using the keypoint co-ordinates andcorresponding image features as inputs enables highly sample-efficientreinforcement learning; (2) learning to explore by controlling keypointlocations drastically reduces the search space, enabling deep exploration(leading to states unreachable through random action exploration) without anyextrinsic rewards.

Unsupervised Learning of Object Keypoints for Perception and Control