SpinalNet: Deep Neural Network with Gradual Input

Open paper

Benchmark	Model	Rank	Results
fine-grained-image-classification-on-bird-225	VGG-19bn (Spinal FC)	#4	Accuracy: 99.02
fine-grained-image-classification-on-bird-225	VGG-19bn	#5	Accuracy: 98.67
fine-grained-image-classification-on-caltech	Wide-ResNet-101 (Spinal FC)	#2	Top-1 Error Rate: 2.68%Accuracy: 97.32
fine-grained-image-classification-on-caltech	Wide-ResNet-101	#3	Top-1 Error Rate: 2.89%
fine-grained-image-classification-on-caltech	VGG-19bn (Spinal FC)	#7	Top-1 Error Rate: 6.84%
fine-grained-image-classification-on-oxford	Wide-ResNet-101 (Spinal FC)	#5	Accuracy: 99.30%
image-classification-on-emnist-balanced	VGG-5(Spinal FC)	#4	Accuracy: 91.05Trainable Parameters: 3630000
image-classification-on-emnist-balanced	VGG-5	#5	Accuracy: 91.04Trainable Parameters: 3646000
image-classification-on-emnist-balanced	CNN(Spinal FC)	#8	Accuracy: 83.21Trainable Parameters: 16050
image-classification-on-emnist-balanced	CNN(Spinal FC)	#9	Accuracy: 82.77Trainable Parameters: 13820
image-classification-on-emnist-balanced	CNN	#10	Accuracy: 79.61Trainable Parameters: 21840
image-classification-on-emnist-digits	VGG-5(Spinal FC)	#3	Accuracy (%): 99.75
image-classification-on-emnist-letters	VGG-5(Spinal FC)	#4	Accuracy: 95.88
image-classification-on-emnist-letters	VGG-5	#5	Accuracy: 95.86
image-classification-on-flowers-102	Wide-ResNet-101 (Spinal FC)	#13	Accuracy: 99.30
image-classification-on-kuzushiji-mnist	VGG-5 (Spinal FC)	#3	Accuracy: 99.15Error: 0.85
image-classification-on-mnist	VGG-5 (Spinal FC)	#13	Percentage error: 0.28Accuracy: 99.72
image-classification-on-stl-10	Wide-ResNet-101 (Spinal FC)	#2	Percentage correct: 98.66
image-classification-on-stl-10	VGG-19bn	#10	Percentage correct: 95.44