Recognizing Electronic Circuits to Enrich Web Documents for Electronic Simulation

Graph-Based Object Detection Enhancement for Symbolic Engineering Drawings

Abstract: The identification of graphic symbols and interconnections is a primary task in the digitization of symbolic engineering diagram images like circuit diagrams. Recent approaches propose the use of Convolutional Neural Networks to the identification of symbols in engineering diagrams. Although recall and precision from CNN based object recognition algorithms are high, false negatives result in some input symbols being missed or misclassified. The missed symbols induce errors in the circuit level features of the extracted circuit, which can be identified using graph level analysis. In this work, a custom annotated printed circuit image set, which is made publicly available in conjunction with the source code of the experiments of this paper, is used to fine-tune a Faster RCNN network to recognise component symbols and blob detection to identify inter-connections between symbols to generate a graph representation of the extracted circuit components. The graph structure is then analysed using graph convolutional neural networks and node degree comparison to identify graph anomalies potentially resulting from false negatives from the object recognition module. Anomaly predictions are then used to identify image regions with potential missed symbols, which are subject to image transforms and re-input to the Faster RCNN, which results in a significant improvement in component recall, which increases to 91% on the test set. The general tools used by the analysis pipeline can also be applied to other Engineering Diagrams with the availability of similar datasets.

LIBSVM: A library for support vector machines

Abstract: LIBSVM is a library for Support Vector Machines (SVMs). We have been actively developing this package since the year 2000. The goal is to help users to easily apply SVM to their applications. LIBSVM has gained wide popularity in machine learning and many other areas. In this article, we present all implementation details of LIBSVM. Issues such as solving SVM optimization problems theoretical convergence multiclass classification probability estimates and parameter selection are discussed in detail.

Histograms of oriented gradients for human detection

Abstract: We study the question of feature sets for robust visual object recognition; adopting linear SVM based human detection as a test case. After reviewing existing edge and gradient based descriptors, we show experimentally that grids of histograms of oriented gradient (HOG) descriptors significantly outperform existing feature sets for human detection. We study the influence of each stage of the computation on performance, concluding that fine-scale gradients, fine orientation binning, relatively coarse spatial binning, and high-quality local contrast normalization in overlapping descriptor blocks are all important for good results. The new approach gives near-perfect separation on the original MIT pedestrian database, so we introduce a more challenging dataset containing over 1800 annotated human images with a large range of pose variations and backgrounds.

Multi-column deep neural networks for image classification

Abstract: Traditional methods of computer vision and machine learning cannot match human performance on tasks such as the recognition of handwritten digits or traffic signs. Our biologically plausible, wide and deep artificial neural network architectures can. Small (often minimal) receptive fields of convolutional winner-take-all neurons yield large network depth, resulting in roughly as many sparsely connected neural layers as found in mammals between retina and visual cortex. Only winner neurons are trained. Several deep neural columns become experts on inputs preprocessed in different ways; their predictions are averaged. Graphics cards allow for fast training. On the very competitive MNIST handwriting benchmark, our method is the first to achieve near-human performance. On a traffic sign recognition benchmark it outperforms humans by a factor of two. We also improve the state-of-the-art on a plethora of common image classification benchmarks.

Probability Estimates for Multi-class Classification by Pairwise Coupling

Abstract: Pairwise coupling is a popular multi-class classification method that combines all comparisons for each pair of classes. This paper presents two approaches for obtaining class probabilities. Both methods can be reduced to linear systems and are easy to implement. We show conceptually and experimentally that the proposed approaches are more stable than the two existing popular methods: voting and the method by Hastie and Tibshirani (1998)

Two decades of statistical language modeling: where do we go from here?

Abstract: Statistical language models estimate the distribution of various natural language phenomena for the purpose of speech recognition and other language technologies. Since the first significant model was proposed in 1980, many attempts have been made to improve the state of the art. We review them, point to a few promising directions, and argue for a Bayesian approach to integration of linguistic theories with data.