AI Paper Atlas

Research map / Learning theory and algorithms / Kernels and regression theory

Tight and Efficient Upper Bound on Spectral Norm of Convolutional Layers

Ekaterina Grishina, Mikhail Gorbunov, Maxim Rakhuba

ECCV 2024 · Learning theory and algorithms › Kernels and regression theory · approach: Generalization theory of deep networks · 3 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
use TN bound (tensor spectral norm) instead of F4 boundspectral norm upper bound · Gaussian kernels, 3x3 to 7x7bound / exact ratio: F4 1.646-2.647 vs TN 1.008-1.131 of exact normImproved
add TN spectral norm regularizationResNet18 · CIFAR100 (with weight decay)top-1 accuracy: 73.84 -> 74.99Improved
add TN spectral norm regularizationResNet18 · CIFAR100 (w/o weight decay)top-1 accuracy: 73.10 -> 73.96Improved
add TN spectral norm regularizationResNet34 · ImageNetAcc@1: 73.368 -> 73.510No clear effect
replace F4 with TN as the regularized boundResNet18 · CIFAR100 (with weight decay)top-1 accuracy: 74.91 -> 74.99No clear effect
TN regularization instead of LipBound for training costResNet18 · CIFAR100 time per epochtime per epoch (s): 242.5 -> 143.6Improved
add 2norm orthogonal regularizer on convolution kernelResNet18 · CIFAR100top-1 accuracy: 73.84 -> 75.99Improved
add Ratio regularizerResNet18 · CIFAR100top-1 accuracy: 73.84 -> 74.98Improved
increase 2norm coefficient from 5e-3 to 1e-2 (ablation)2norm regularizer · CIFAR100top-1 accuracy: 75.99 -> 74.43Hurt
2norm vs OCNN orthogonal regularizer in costResNet18 · CIFAR100time per epoch (s): 150.1 -> 171.8Hurt

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