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

We present a single neural network architecture composed of task-agnosticcomponents (ViTs, convolutions, and LSTMs) that achieves state-of-art resultson both the ImageNav ("go to location in <this picture>") and ObjectNav ("finda chair") tasks without any task-specific modules like object detection,segmentation, mapping, or planning modules. Such general-purpose methods offeradvantages of simplicity in design, positive scaling with available compute,and versatile applicability to multiple tasks. Our work builds upon the recentsuccess of self-supervised learning (SSL) for pre-training vision transformers(ViT). However, while the training recipes for convolutional networks aremature and robust, the recipes for ViTs are contingent and brittle, and in thecase of ViTs for visual navigation, yet to be fully discovered. Specifically,we find that vanilla ViTs do not outperform ResNets on visual navigation. Wepropose the use of a compression layer operating over ViT patch representationsto preserve spatial information along with policy training improvements. Theseimprovements allow us to demonstrate positive scaling laws for the first timein visual navigation tasks. Consequently, our model advances state-of-the-artperformance on ImageNav from 54.2% to 82.0% success and performs competitivelyagainst concurrent state-of-art on ObjectNav with success rate of 64.0% vs.65.0%. Overall, this work does not present a fundamentally new approach, butrather recommendations for training a general-purpose architecture thatachieves state-of-art performance today and could serve as a strong baselinefor future methods.

OVRL-V2: A simple state-of-art baseline for ImageNav and ObjectNav