Some Improvements on Deep Convolutional Neural Network Based Image Classification

By Andrew G. Howard. 2013

443 citations recorded by the source corpus.

Abstract: We investigate multiple techniques to improve upon the current state of the art deep convolutional neural network based image classification pipeline. The techiques include adding more image transformations to training data, adding more transformations to generate additional predictions at test time and using complementary models applied to higher resolution images. This paper summarizes our entry in the Imagenet Large Scale Visual Recognition Challenge 2013. Our system achieved a top 5 classification error rate of 13.55% using no external data which is over a 20% relative improvement on the previous year's winner.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Computer Vision and Pattern Recognition

Citations in the corpus

Delving Deep into Rectifiers: Surpassing Human-Level Performance on ImageNet Classification (2015, background): “techniques [13, 30, 10, 36], aggressive data augmentation [18, 14, 29, 33], and large-scale data [4, 26].”
Convolutional neural networks at constrained time cost (2014, methodology): “No other data augmentation, such as scale jittering [10, 22, 9], is used.”
Going deeper with convolutions (2014, methodology): “Also, we found that the photometric distortions of Andrew Howard [8] were useful to combat overfitting to the imaging conditions of training data.”
Very Deep Convolutional Networks for Large-Scale Image Recognition (2014, background): “Another line of improvements dealt with training and testing the networks densely over the whol e image and over multiple scales [7, 16].”
ImageNet Large Scale Visual Recognition Challenge (2014, methodology): “, 2014a) incorporating improvements of (Howard, 2014).”
Spatial Pyramid Pooling in Deep Convolutional Networks for Visual Recognition (2014, methodology): “Next we describe our training solution that takes advantage of these GPU implementations while still preserving the spatial pyramid pooling behaviors.”
A Simple Framework for Contrastive Learning of Visual Representations (2020, neutral): “The other type of augmentation involves appearance transformation, such as color distortion (including color dropping, brightness, contrast, saturation, hue) (Howard, 2013; Szegedy et al., 2015), Gaussian blur, and Sobel ﬁltering.”
SSD: Single Shot MultiBox Detector (2015, methodology): “5, in addition to applying some photometric distortions similar to those described in [10].”
Pix2seq: A Language Modeling Framework for Object Detection (2021, methodology): “Following (Howard, 2013; Chen et al., 2020a;b), we also use color distortion with a strength of 0.5.”
AutoAugment: Learning Augmentation Policies from Data (2018, methodology): “For baseline augmentation, we use the standard Inception-style pre-processing which involves scaling pixel values to [-1,1], horizontal ﬂips with 50% probability, and random distortions of colors [14, 67].”
VoxelNet: End-to-End Learning for Point Cloud Based 3D Object Detection (2017, background): “Introducing global scale augmentation improves robustness of the network for detecting objects with various sizes and distances as shown in image-based classification [37, 18] and detection tasks [12, 17].”
Understanding Data Augmentation for Classification: When to Warp? (2016, background): “A good overview of label preserving transformations for image data for a more modern classficati on architecture is given by [17].”
Efficient and accurate approximations of nonlinear convolutional networks (2014, methodology): “Besides the standard strategies used in [11], we train this model using the 224×224 views cropped from resized images whose shorter edge is 256 [9].”
Dilated Residual Networks (2017, methodology): “We use scale and aspect ratio augmentation as in Szegedy et al. [14] and color perturbation as in Krizhevsky et al. [8] and Howard [7].”
Deep Image: Scaling up Image Recognition (2015, methodology): “We use data augmentation in novel ways, much more aggressively than previous work (Howard, 2013), (Krizhevsky et al., 2012).”