Proceedings Article•
ImageNet Classification with Deep Convolutional Neural Networks
03 Dec 2012-Vol. 25, pp 1097-1105
TL;DR: The state-of-the-art performance of CNNs was achieved by Deep Convolutional Neural Networks (DCNNs) as discussed by the authors, which consists of five convolutional layers, some of which are followed by max-pooling layers, and three fully-connected layers with a final 1000-way softmax.
Abstract: We trained a large, deep convolutional neural network to classify the 1.2 million high-resolution images in the ImageNet LSVRC-2010 contest into the 1000 different classes. On the test data, we achieved top-1 and top-5 error rates of 37.5% and 17.0% which is considerably better than the previous state-of-the-art. The neural network, which has 60 million parameters and 650,000 neurons, consists of five convolutional layers, some of which are followed by max-pooling layers, and three fully-connected layers with a final 1000-way softmax. To make training faster, we used non-saturating neurons and a very efficient GPU implementation of the convolution operation. To reduce overriding in the fully-connected layers we employed a recently-developed regularization method called "dropout" that proved to be very effective. We also entered a variant of this model in the ILSVRC-2012 competition and achieved a winning top-5 test error rate of 15.3%, compared to 26.2% achieved by the second-best entry.
Citations
More filters
Posted Content•
TL;DR: In this article, the authors leverage the natural synchronization between vision and sound to learn an acoustic representation using two-million unlabeled videos and propose a student-teacher training procedure which transfers discriminative visual knowledge from well established visual recognition models into the sound modality using unlabelled video as a bridge.
Abstract: We learn rich natural sound representations by capitalizing on large amounts of unlabeled sound data collected in the wild We leverage the natural synchronization between vision and sound to learn an acoustic representation using two-million unlabeled videos Unlabeled video has the advantage that it can be economically acquired at massive scales, yet contains useful signals about natural sound We propose a student-teacher training procedure which transfers discriminative visual knowledge from well established visual recognition models into the sound modality using unlabeled video as a bridge Our sound representation yields significant performance improvements over the state-of-the-art results on standard benchmarks for acoustic scene/object classification Visualizations suggest some high-level semantics automatically emerge in the sound network, even though it is trained without ground truth labels
725 citations
TL;DR: Experimental results on Pascal VOC 2007 and VOC 2012 multi-label image datasets well demonstrate the superiority of the proposed HCP infrastructure over other state-of-the-arts, where an arbitrary number of object segment hypotheses are taken as the inputs.
Abstract: Convolutional Neural Network (CNN) has demonstrated promising performance in single-label image classification tasks. However, how CNN best copes with multi-label images still remains an open problem, mainly due to the complex underlying object layouts and insufficient multi-label training images. In this work, we propose a flexible deep CNN infrastructure, called Hypotheses-CNN-Pooling (HCP), where an arbitrary number of object segment hypotheses are taken as the inputs, then a shared CNN is connected with each hypothesis, and finally the CNN output results from different hypotheses are aggregated with max pooling to produce the ultimate multi-label predictions. Some unique characteristics of this flexible deep CNN infrastructure include: 1) no ground-truth bounding box information is required for training; 2) the whole HCP infrastructure is robust to possibly noisy and/or redundant hypotheses; 3) the shared CNN is flexible and can be well pre-trained with a large-scale single-label image dataset, e.g., ImageNet; and 4) it may naturally output multi-label prediction results. Experimental results on Pascal VOC 2007 and VOC 2012 multi-label image datasets well demonstrate the superiority of the proposed HCP infrastructure over other state-of-the-arts. In particular, the mAP reaches 90.5% by HCP only and 93.2% after the fusion with our complementary result in [12] based on hand-crafted features on the VOC 2012 dataset.
722 citations
TL;DR: The Siamese U-Net outperforms current building extraction methods and could provide valuable reference and the designed experiments indicate the data set is accurate and can serve multiple purposes including building instance segmentation and change detection.
Abstract: The application of the convolutional neural network has shown to greatly improve the accuracy of building extraction from remote sensing imagery. In this paper, we created and made open a high-quality multisource data set for building detection, evaluated the accuracy obtained in most recent studies on the data set, demonstrated the use of our data set, and proposed a Siamese fully convolutional network model that obtained better segmentation accuracy. The building data set that we created contains not only aerial images but also satellite images covering 1000 km2 with both raster labels and vector maps. The accuracy of applying the same methodology to our aerial data set outperformed several other open building data sets. On the aerial data set, we gave a thorough evaluation and comparison of most recent deep learning-based methods, and proposed a Siamese U-Net with shared weights in two branches, and original images and their down-sampled counterparts as inputs, which significantly improves the segmentation accuracy, especially for large buildings. For multisource building extraction, the generalization ability is further evaluated and extended by applying a radiometric augmentation strategy to transfer pretrained models on the aerial data set to the satellite data set. The designed experiments indicate our data set is accurate and can serve multiple purposes including building instance segmentation and change detection; our result shows the Siamese U-Net outperforms current building extraction methods and could provide valuable reference.
721 citations
24 Jul 2016
TL;DR: This method aims to allow using the high-resolution histopathological images from BreaKHis as input to existing CNN, avoiding adaptations of the model that can lead to a more complex and computationally costly architecture.
Abstract: The performance of most conventional classification systems relies on appropriate data representation and much of the efforts are dedicated to feature engineering, a difficult and time-consuming process that uses prior expert domain knowledge of the data to create useful features. On the other hand, deep learning can extract and organize the discriminative information from the data, not requiring the design of feature extractors by a domain expert. Convolutional Neural Networks (CNNs) are a particular type of deep, feedforward network that have gained attention from research community and industry, achieving empirical successes in tasks such as speech recognition, signal processing, object recognition, natural language processing and transfer learning. In this paper, we conduct some preliminary experiments using the deep learning approach to classify breast cancer histopathological images from BreaKHis, a publicly dataset available at http://web.inf.ufpr.br/vri/breast-cancer-database. We propose a method based on the extraction of image patches for training the CNN and the combination of these patches for final classification. This method aims to allow using the high-resolution histopathological images from BreaKHis as input to existing CNN, avoiding adaptations of the model that can lead to a more complex and computationally costly architecture. The CNN performance is better when compared to previously reported results obtained by other machine learning models trained with hand-crafted textural descriptors. Finally, we also investigate the combination of different CNNs using simple fusion rules, achieving some improvement in recognition rates.
720 citations
Posted Content•
TL;DR: This work revisits both retrieval stages, namely initial search and re-ranking, by employing the same primitive information derived from the CNN, and significantly improves existing CNN-based recognition pipeline.
Abstract: Recently, image representation built upon Convolutional Neural Network (CNN) has been shown to provide effective descriptors for image search, outperforming pre-CNN features as short-vector representations. Yet such models are not compatible with geometry-aware re-ranking methods and still outperformed, on some particular object retrieval benchmarks, by traditional image search systems relying on precise descriptor matching, geometric re-ranking, or query expansion. This work revisits both retrieval stages, namely initial search and re-ranking, by employing the same primitive information derived from the CNN. We build compact feature vectors that encode several image regions without the need to feed multiple inputs to the network. Furthermore, we extend integral images to handle max-pooling on convolutional layer activations, allowing us to efficiently localize matching objects. The resulting bounding box is finally used for image re-ranking. As a result, this paper significantly improves existing CNN-based recognition pipeline: We report for the first time results competing with traditional methods on the challenging Oxford5k and Paris6k datasets.
719 citations
References
More filters
[...]
TL;DR: Internal estimates monitor error, strength, and correlation and these are used to show the response to increasing the number of features used in the forest, and are also applicable to regression.
Abstract: Random forests are a combination of tree predictors such that each tree depends on the values of a random vector sampled independently and with the same distribution for all trees in the forest. The generalization error for forests converges a.s. to a limit as the number of trees in the forest becomes large. The generalization error of a forest of tree classifiers depends on the strength of the individual trees in the forest and the correlation between them. Using a random selection of features to split each node yields error rates that compare favorably to Adaboost (Y. Freund & R. Schapire, Machine Learning: Proceedings of the Thirteenth International conference, aaa, 148–156), but are more robust with respect to noise. Internal estimates monitor error, strength, and correlation and these are used to show the response to increasing the number of features used in the splitting. Internal estimates are also used to measure variable importance. These ideas are also applicable to regression.
79,257 citations
20 Jun 2009
TL;DR: A new database called “ImageNet” is introduced, a large-scale ontology of images built upon the backbone of the WordNet structure, much larger in scale and diversity and much more accurate than the current image datasets.
Abstract: The explosion of image data on the Internet has the potential to foster more sophisticated and robust models and algorithms to index, retrieve, organize and interact with images and multimedia data. But exactly how such data can be harnessed and organized remains a critical problem. We introduce here a new database called “ImageNet”, a large-scale ontology of images built upon the backbone of the WordNet structure. ImageNet aims to populate the majority of the 80,000 synsets of WordNet with an average of 500-1000 clean and full resolution images. This will result in tens of millions of annotated images organized by the semantic hierarchy of WordNet. This paper offers a detailed analysis of ImageNet in its current state: 12 subtrees with 5247 synsets and 3.2 million images in total. We show that ImageNet is much larger in scale and diversity and much more accurate than the current image datasets. Constructing such a large-scale database is a challenging task. We describe the data collection scheme with Amazon Mechanical Turk. Lastly, we illustrate the usefulness of ImageNet through three simple applications in object recognition, image classification and automatic object clustering. We hope that the scale, accuracy, diversity and hierarchical structure of ImageNet can offer unparalleled opportunities to researchers in the computer vision community and beyond.
49,639 citations
01 Jan 1988
TL;DR: This chapter contains sections titled: The Problem, The Generalized Delta Rule, Simulation Results, Some Further Generalizations, Conclusion.
Abstract: This chapter contains sections titled: The Problem, The Generalized Delta Rule, Simulation Results, Some Further Generalizations, Conclusion
17,604 citations
Dissertation•
01 Jan 2009TL;DR: In this paper, the authors describe how to train a multi-layer generative model of natural images, using a dataset of millions of tiny colour images, described in the next section.
Abstract: In this work we describe how to train a multi-layer generative model of natural images. We use a dataset of millions of tiny colour images, described in the next section. This has been attempted by several groups but without success. The models on which we focus are RBMs (Restricted Boltzmann Machines) and DBNs (Deep Belief Networks). These models learn interesting-looking filters, which we show are more useful to a classifier than the raw pixels. We train the classifier on a labeled subset that we have collected and call the CIFAR-10 dataset.
15,005 citations
Proceedings Article•
21 Jun 2010TL;DR: Restricted Boltzmann machines were developed using binary stochastic hidden units that learn features that are better for object recognition on the NORB dataset and face verification on the Labeled Faces in the Wild dataset.
Abstract: Restricted Boltzmann machines were developed using binary stochastic hidden units. These can be generalized by replacing each binary unit by an infinite number of copies that all have the same weights but have progressively more negative biases. The learning and inference rules for these "Stepped Sigmoid Units" are unchanged. They can be approximated efficiently by noisy, rectified linear units. Compared with binary units, these units learn features that are better for object recognition on the NORB dataset and face verification on the Labeled Faces in the Wild dataset. Unlike binary units, rectified linear units preserve information about relative intensities as information travels through multiple layers of feature detectors.
14,799 citations