Abstract | ||
---|---|---|
In this brief paper, we propose a method of feature extraction for digit recognition that is inspired by vision research: a sparse-coding strategy and a local maximum operation. We show that our method, despite its simplicity, yields state-of-the-art classification results on a highly competitive digit-recognition benchmark. We first employ the unsupervised Sparsenet algorithm to learn a basis for representing patches of handwritten digit images. We then use this basis to extract local coefficients. In a second step, we apply a local maximum operation to implement local shift invariance. Finally, we train a support vector machine (SVM) on the resulting feature vectors and obtain state-of-the-art classification performance in the digit recognition task defined by the MNIST benchmark. We compare the different classification performances obtained with sparse coding, Gabor wavelets, and principal component analysis (PCA). We conclude that the learning of a sparse representation of local image patches combined with a local maximum operation for feature extraction can significantly improve recognition performance. |
Year | DOI | Venue |
---|---|---|
2008 | 10.1109/TNN.2008.2005830 | IEEE Transactions on Neural Networks |
Keywords | Field | DocType |
recognition performance,different classification performance,simple method,handwritten digit image,sparse coding,local image,digit recognition task,svm,high-performance digit recognition,feature extraction,local coefficient,local maximum operation,local shift invariance,digit recognition,digital image,sparse representation,principle component analysis,gabor wavelet,shift invariant,image classification,principal component analysis,support vector machines,support vector machine,gabor wavelets,wavelet transforms,feature vector,classification algorithms | MNIST database,Gabor wavelet,Computer science,Gabor filter,Artificial intelligence,Contextual image classification,Feature vector,Pattern recognition,Sparse approximation,Speech recognition,Feature extraction,Statistical classification,Machine learning | Journal |
Volume | Issue | ISSN |
19 | 11 | 1941-0093 |
Citations | PageRank | References |
46 | 2.07 | 17 |
Authors | ||
3 |
Name | Order | Citations | PageRank |
---|---|---|---|
Kai Labusch | 1 | 113 | 8.50 |
Erhardt Barth | 2 | 653 | 58.33 |
Thomas Martinetz | 3 | 65 | 6.13 |