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 propose in this paper a Quantized Distilled Low-Rank Neural Radiance Field(QDLR-NeRF) representation for the task of light field compression. Whileexisting compression methods encode the set of light field sub-aperture images,our proposed method learns an implicit scene representation in the form of aNeural Radiance Field (NeRF), which also enables view synthesis. To reduce itssize, the model is first learned under a Low-Rank (LR) constraint using aTensor Train (TT) decomposition within an Alternating Direction Method ofMultipliers (ADMM) optimization framework. To further reduce the model's size,the components of the tensor train decomposition need to be quantized. However,simultaneously considering the optimization of the NeRF model with both thelow-rank constraint and rate-constrained weight quantization is challenging. Toaddress this difficulty, we introduce a network distillation operation thatseparates the low-rank approximation and the weight quantization during networktraining. The information from the initial LR-constrained NeRF (LR-NeRF) isdistilled into a model of much smaller dimension (DLR-NeRF) based on the TTdecomposition of the LR-NeRF. We then learn an optimized global codebook toquantize all TT components, producing the final QDLR-NeRF. Experimental resultsshow that our proposed method yields better compression efficiency compared tostate-of-the-art methods, and it additionally has the advantage of allowing thesynthesis of any light field view with high quality.

Distilled Low Rank Neural Radiance Field with Quantization for Light Field Compression