• Corpus ID: 13439545

Analysis and Optimization of Convolutional Neural Network Architectures

  title={Analysis and Optimization of Convolutional Neural Network Architectures},
  author={Martin Thoma},
Convolutional Neural Networks (CNNs) dominate various computer vision tasks since Alex Krizhevsky showed that they can be trained effectively and reduced the top-5 error from 26.2 % to 15.3 % on the ImageNet large scale visual recognition challenge. [] Key Method A novel way to visualize classification errors with confusion matrices was developed. Based on this method, hierarchical classifiers are described and evaluated. Additionally, some results are confirmed and quantified for CIFAR-100. For example, the…

Investigating the Robustness of Pre-trained Networks on OCT-Dataset

This work investigates the effectiveness and the application of pre-trained models from natural (non-medical) images to images from the OCT (optical coherence tomography) domain in ophthalmology and shows the robustness of a series of models without the demand to train a model from scratch again.

Broadcasting Convolutional Network

This paper proposes Broadcasting Convolutional Networks (BCN) that allow global receptive fields to share spatial information, and utilizes BCN to propose Multi-Relational Networks (multiRN) that greatly improve existing Relation Networks (RNs).

Faster Neural Network Training with Approximate Tensor Operations

A novel technique for faster Neural Network (NN) training by systematically approximating all the constituent matrix multiplications and convolutions, complementary to other approximation techniques, requires no changes to the dimensions of the network layers, and is compatible with existing training frameworks.

Surface Defect Detection for Automated Inspection Systems using Convolutional Neural Networks

Custom CNNs and a transfer-learned AlexNet are applied to an experimental data set with artificial defects in order to analyze suitability and required network depth for surface inspections, yielding a classification accuracy of up to 99 % with a single CNN.

Visualizing CoAtNet Predictions for Aiding Melanoma Detection

  • Daniel Kvak
  • Computer Science
    Engineering and Technology Journal
  • 2022
This paper proposes a multi-class classification task using the CoAtNet architecture, a hybrid model that combines the deep depthwise convolution matrix operation of traditional convolutional neural networks with the strengths of Transformer models and self-attention mechanics to achieve better generalization and capacity.

Exploring Convolutional Neural Networks on the ρ-VEX architecture

This work developed a streaming pipeline in a simulator designed to execute CNN inference and take advantage of the overlapped execution to increase throughput and proposes several optimizations to increase performance of CNN inference on the ρ-VEX architecture.

Broadcasting Convolutional Network for Visual Relational Reasoning

The Broadcasting Convolutional Network (BCN) is proposed that extracts key object features from the global field of an entire input image and recognizes their relationship with local features and the Multi-Relational Network (multiRN) that improves the existing Relation Network (RN) by utilizing the BCN module is introduced.

Multiclass Image Classification Using GANs and CNN Based on Holes Drilled in Laminated Chipboard

The aim of the research was to create a model capable of identifying different levels of quality of the holes, where the reduced quality would serve as a warning that the drill is about to wear down, which could reduce the damage caused by a blunt tool.

Learning to Segment Vessels from Poorly Illuminated Fundus Images

This paper analyses the performance of U-Net architecture on DRIVE and RIM-ONE datasets, and a different approach for data augmentation using vignetting masks is presented to create more annotated fundus data.



Doubly Convolutional Neural Networks

This paper proposes doubly convolutional neural networks (DCNNs), which significantly improve the performance of CNNs by further exploring this idea and shows that DCNN can serve the dual purpose of building more accurate models and/or reducing the memory footprint without sacrificing the accuracy.

ImageNet classification with deep convolutional neural networks

A large, deep convolutional neural network was trained to classify the 1.2 million high-resolution images in the ImageNet LSVRC-2010 contest into the 1000 different classes and employed a recently developed regularization method called "dropout" that proved to be very effective.

Visualizing and Understanding Convolutional Networks

A novel visualization technique is introduced that gives insight into the function of intermediate feature layers and the operation of the classifier in large Convolutional Network models, used in a diagnostic role to find model architectures that outperform Krizhevsky et al on the ImageNet classification benchmark.

Rethinking the Inception Architecture for Computer Vision

This work is exploring ways to scale up networks in ways that aim at utilizing the added computation as efficiently as possible by suitably factorized convolutions and aggressive regularization.

Genetic CNN

  • Lingxi XieA. Yuille
  • Computer Science
    2017 IEEE International Conference on Computer Vision (ICCV)
  • 2017
The core idea is to propose an encoding method to represent each network structure in a fixed-length binary string to efficiently explore this large search space.

Visualizing Deep Convolutional Neural Networks Using Natural Pre-images

This paper studies several landmark representations, both shallow and deep, by a number of complementary visualization techniques based on the concept of “natural pre-image”, and shows that several layers in CNNs retain photographically accurate information about the image, with different degrees of geometric and photometric invariance.

Very Deep Convolutional Networks for Large-Scale Image Recognition

This work investigates the effect of the convolutional network depth on its accuracy in the large-scale image recognition setting using an architecture with very small convolution filters, which shows that a significant improvement on the prior-art configurations can be achieved by pushing the depth to 16-19 weight layers.

Densely Connected Convolutional Networks

The Dense Convolutional Network (DenseNet), which connects each layer to every other layer in a feed-forward fashion, and has several compelling advantages: they alleviate the vanishing-gradient problem, strengthen feature propagation, encourage feature reuse, and substantially reduce the number of parameters.

Some Improvements on Deep Convolutional Neural Network Based Image Classification

This paper summarizes the entry in the Imagenet Large Scale Visual Recognition Challenge 2013, which achieved a top 5 classification error rate and achieved over a 20% relative improvement on the previous year's winner.