AI Paper Atlas

Research map / Learning theory and algorithms / Matrix and tensor methods

Trading Positional Complexity vs Deepness in Coordinate Networks

Jianqiao Zheng, Sameera Ramasinghe, Xueqian Li, Simon Lucey

ECCV 2022 · Learning theory and algorithms › Matrix and tensor methods · approach: Geometry of low-rank and SPD structure · 13 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
replace simple encoding + MLP with complex Kronecker encoding + single linear layer (Gau)Gaussian embedder · 2D image, separablePSNR (dB): 26.17 -> 26.69Improved
replace simple encoding + MLP with complex encoding (Tri)Triangle embedder · 2D image, separablePSNR (dB): 25.96 -> 26.65Improved
use complex encoding instead of simple encoding for frequency featuresLinF · 2D image, separablePSNR (dB): 26.12 -> 19.73Hurt
use complex encoding with a single linear layerRFF · 2D image, separablePSNR (dB): 26.58 -> 24.34Hurt
solve complex encoding in closed form instead of gradient descentGaussian embedder · 2D image, separabletraining time (s): 0.18Improved
use complex encoding instead of simple encoding (training time)Gaussian embedder · 2D image, separabletraining time (s): 23.12 -> 2.01Improved
use complex encoding with a single linear layerGaussian embedder · 2D image, random 25%PSNR (dB): 24.63 -> 23.97Hurt
use complex encoding with a single linear layerTriangle embedder · 2D image, random 25%PSNR (dB): 24.24 -> 23.21Hurt
use complex encoding with a single linear layerTriangle embedder · videoPSNR (dB): 20.86 -> 22.24Improved
use complex encoding with a single linear layerLinF · videoPSNR (dB): 21.49 -> 9.76Hurt
solve complex encoding in closed formTriangle embedder · videotraining time (s): 78.91 -> 0.009Improved
add one hidden layer to a linear network when embedding rank is low (ablation)Gaussian embedder · 1D signalreconstruction performanceImproved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy