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 visual signal compression is a long-standing problem. Fueled by therecent advances of deep learning, exciting progress has been made. Despitebetter compression performance, existing end-to-end compression algorithms arestill designed towards better signal quality in terms of rate-distortionoptimization. In this paper, we show that the design and optimization ofnetwork architecture could be further improved for compression towards machinevision. We propose an inverted bottleneck structure for the encoder of theend-to-end compression towards machine vision, which specifically accounts forefficient representation of the semantic information. Moreover, we quest thecapability of optimization by incorporating the analytics accuracy into theoptimization process, and the optimality is further explored with generalizedrate-accuracy optimization in an iterative manner. We use object detection as ashowcase for end-to-end compression towards machine vision, and extensiveexperiments show that the proposed scheme achieves significant BD-rate savingsin terms of analysis performance. Moreover, the promise of the scheme is alsodemonstrated with strong generalization capability towards other machine visiontasks, due to the enabling of signal-level reconstruction.

End-to-end Compression Towards Machine Vision: Network Architecture Design and Optimization