SpinalNet: Deep Neural Network with Gradual Input

arXiv 2020 H M Dipu KabirMoloud AbdarSeyed Mohammad Jafar JalaliAbbas KhosraviAmir F AtiyaSaeid NahavandiDipti Srinivasan

Over the past few years, deep neural networks (DNNs) have garnered remarkable success in a diverse range of real-world applications. However, DNNs consider a large number of inputs and consist of a large number of parameters, resulting in high computational demand... (read more)

PDF Abstract

Results from the Paper


TASK DATASET MODEL METRIC NAME METRIC VALUE GLOBAL RANK USES EXTRA
TRAINING DATA
RESULT BENCHMARK
Fine-Grained Image Classification Bird-225 Wide-ResNet-101 (Spinal FC) Accuracy 99.56 # 1
Fine-Grained Image Classification Bird-225 Wide-ResNet-101 Accuracy 99.38 # 2
Fine-Grained Image Classification Bird-225 VGG-19bn Accuracy 98.67 # 4
Fine-Grained Image Classification Bird-225 VGG-19bn (Spinal FC) Accuracy 99.02 # 3
Fine-Grained Image Classification Caltech-101 Wide-ResNet-101 (Spinal FC) Top-1 Error Rate 2.68% # 1
Fine-Grained Image Classification Caltech-101 VGG-19bn Top-1 Error Rate 7.02% # 4
Fine-Grained Image Classification Caltech-101 VGG-19bn (Spinal FC) Top-1 Error Rate 6.84% # 3
Fine-Grained Image Classification Caltech-101 Wide-ResNet-101 Top-1 Error Rate 2.89% # 2
Image Classification CIFAR-10 VGG-19(Spinal FC) Percentage correct 91.40 # 63
Percentage error 8.60 # 29
Image Classification CIFAR-10 Wide-ResNet-101 Percentage correct 98.22 # 10
Image Classification CIFAR-100 VGG-16(Spinal FC) Percentage correct 64.99 # 65
Percentage error 35.01 # 24
Image Classification CIFAR-100 Wide-ResNet-101 (Spinal FC) Percentage correct 88.34 # 7
Image Classification EMNIST-Balanced VGG-5 Accuracy 91.04 # 2
Image Classification EMNIST-Balanced VGG-5(Spinal FC) Accuracy 91.05 # 1
Image Classification EMNIST-Digits VGG-5 Accuracy (%) 99.81 # 2
Image Classification EMNIST-Digits VGG-5(Spinal FC) Accuracy (%) 99.82 # 1
Image Classification EMNIST-Letters VGG-5(Spinal FC) Accuracy 95.88 # 1
Image Classification EMNIST-Letters VGG-5 Accuracy 95.86 # 2
Image Classification Fashion-MNIST VGG-5(Spinal FC) Percentage error 5.32 # 6
Image Classification Flowers-102 Wide-ResNet-101 (Spinal FC) Accuracy 99.30% # 3
Image Classification Flowers-102 Wide-ResNet-101 Accuracy 99.39% # 2
Fine-Grained Image Classification Fruits-360 Wide-ResNet-101 (Spinal FC) Accuracy (%) 100 # 1
Fine-Grained Image Classification Fruits-360 Wide-ResNet-101 Accuracy (%) 99.96 # 2
Fine-Grained Image Classification Fruits-360 VGG-19bn (Spinal FC) Accuracy (%) 99.96 # 2
Fine-Grained Image Classification Fruits-360 VGG-19bn Accuracy (%) 99.90 # 3
Image Classification Kuzushiji-MNIST VGG-5 Accuracy 98.94 # 4
Image Classification Kuzushiji-MNIST VGG-5 (Spinal FC) Accuracy 99.15 # 1
Error 0.85 # 1
Image Classification MNIST VGG-5 Accuracy 99.72 # 3
Image Classification MNIST VGG-5 (Spinal FC) Percentage error 0.28 # 9
Accuracy 99.72 # 3
Fine-Grained Image Classification Oxford 102 Flowers Wide-ResNet-101 (Spinal FC) Accuracy 99.30% # 3
Fine-Grained Image Classification Oxford 102 Flowers Wide-ResNet-101 Accuracy 99.39% # 2
Image Classification QMNIST VGG-5(Spinal FC) Accuracy (%) 99.68 # 1
Image Classification STL-10 VGG-19bn Percentage correct 95.44 # 10
Image Classification STL-10 Wide-ResNet-101 (Spinal FC) Percentage correct 98.66 # 1
Image Classification STL-10 Wide-ResNet-101 Percentage correct 98.40 # 2
Image Classification STL-10 VGG-19bn (Spinal FC) Percentage correct 95.57 # 8

Methods used in the Paper


METHOD TYPE
🤖 No Methods Found Help the community by adding them if they're not listed; e.g. Deep Residual Learning for Image Recognition uses ResNet