North American Chapter of the Association for Computational Linguistics, 2005
Key phrases are usually among the most information-bearing linguistic structures. Translating the... more Key phrases are usually among the most information-bearing linguistic structures. Translating them correctly will improve many natural language processing applications. We propose a new framework to mine key phrase translations from web corpora. We submit a source phrase to a search engine as a query, then expand queries by adding the translations of topic-relevant hint words from the returned snippets. We retrieve mixedlanguage web pages based on the expanded queries. Finally, we extract the key phrase translation from the secondround returned web page snippets with phonetic, semantic and frequencydistance features. We achieve 46% phrase translation accuracy when using top 10 returned snippets, and 80% accuracy with 165 snippets. Both results are significantly better than several existing methods.
In this paper we describe the components of our statistical machine translation system. This syst... more In this paper we describe the components of our statistical machine translation system. This system combines phrase-tophrase translations extracted from a bilingual corpus using different alignment approaches. Special methods to extract and align named entities are used. We show how a manual lexicon can be incorporated into the statistical system in an optimized way. Experiments on Chinese-to-English and Arabic-to-English translation tasks are presented.
Energy-Based Models (EBMs) capture dependencies between variables by associating a scalar energy ... more Energy-Based Models (EBMs) capture dependencies between variables by associating a scalar energy to each configuration of the variables. Inference consists in clamping the value of observed variables and finding configurations of the remaining variables that minimize the energy. Learning consists in finding an energy function in which observed configurations of the variables are given lower energies than unobserved ones. The EBM approach provides a common theoretical framework for many learning models, including traditional discriminative and generative approaches, as well as graph-transformer networks, conditional random fields, maximum margin Markov networks, and several manifold learning methods.
Probabilistic graphical models associate a probability to each configuration of the relevant vari... more Probabilistic graphical models associate a probability to each configuration of the relevant variables. Energy-based models (EBM) associate an energy to those configurations, eliminating the need for proper normalization of probability distributions. Making a decision (an inference) with an EBM consists in comparing the energies associated with various configurations of the variable to be predicted, and choosing the one with the smallest energy. Such systems must be trained discriminatively to associate low energies to the desired configurations and higher energies to undesired configurations. A wide variety of loss function can be used for this purpose. We give sufficient conditions that a loss function should satisfy so that its minimization will cause the system to approach to desired behavior. We give many specific examples of suitable loss functions, and show an application to object recognition in images. it is important to note that the energy is quantity minimized
International Conference on Multimedia Computing and Systems/International Conference on Multimedia and Expo, 2000
Describes a real-time face-tracking algorithm. We start with single face tracking based on statis... more Describes a real-time face-tracking algorithm. We start with single face tracking based on statistical color modeling and a deformable template. We then expand the algorithm to track multiple faces, possibly with occlusion, by constraining the speed and size changes of the faces. We test the algorithm on sequences with different occlusion patterns and analyze the tracking performance. We also present a tracking software library based on this algorithm. This library can be applied to human-computer interfaces, lip-reading and virtual environments
We present an unsupervised method for learning a hierarchy of sparse feature detectors that are i... more We present an unsupervised method for learning a hierarchy of sparse feature detectors that are invariant to small shifts and distortions. The resulting feature extractor consists of multiple convolution filters, followed by a feature-pooling layer that computes the max of each filter output within adjacent windows, and a point-wise sigmoid non-linearity. A second level of larger and more invariant features is obtained by training the same algorithm on patches of features from the first level. Training a supervised classifier on these features yields 0.64% error on MNIST, and 54% average recognition rate on Caltech 101 with 30 training samples per category. While the resulting architecture is similar to convolutional networks, the layer-wise unsupervised training procedure alleviates the over-parameterization problems that plague purely supervised learning procedures, and yields good performance with very few labeled training samples.
We assess the applicability of several popular learning methods for the problem of recognizing ge... more We assess the applicability of several popular learning methods for the problem of recognizing generic visual categories with invariance to pose, lighting, and surrounding clutter. A large dataset comprising stereo image pairs of 50 uniform-colored toys under 36 azimuths, 9 elevations, and 6 lighting conditions was collected (for a total of 194,400 individual images). The objects were 10 instances of 5 generic categories: four-legged animals, human figures, airplanes, trucks, and cars. Five instances of each category were used for training, and the other five for testing. Low-resolution grayscale images of the objects with various amounts of variability and surrounding clutter were used for training and testing. Nearest neighbor methods, support vector machines, and convolutional networks, operating on raw pixels or on PCA-derived features were tested. Test error rates for unseen object instances placed on uniform backgrounds were around 13% for SVM and 7% for convolutional nets. On a segmentation/recognition task with highly cluttered images, SVM proved impractical, while convolutional nets yielded 16/7% error. A real-time version of the system was implemented that can detect and classify objects in natural scenes at around 10 frames per second.
The detection and recognition of generic object categories with invariance to viewpoint, illumina... more The detection and recognition of generic object categories with invariance to viewpoint, illumination, and clutter requires the combination of a feature extractor and a classifier. We show that architectures such as convolutional networks are good at learning invariant features, but not always optimal for classification, while Support Vector Machines are good at producing decision surfaces from wellbehaved feature vectors, but cannot learn complicated invariances. We present a hybrid system where a convolutional network is trained to detect and recognize generic objects, and a Gaussian-kernel SVM is trained from the features learned by the convolutional network.
North American Chapter of the Association for Computational Linguistics, 2005
Key phrases are usually among the most information-bearing linguistic structures. Translating the... more Key phrases are usually among the most information-bearing linguistic structures. Translating them correctly will improve many natural language processing applications. We propose a new framework to mine key phrase translations from web corpora. We submit a source phrase to a search engine as a query, then expand queries by adding the translations of topic-relevant hint words from the returned snippets. We retrieve mixedlanguage web pages based on the expanded queries. Finally, we extract the key phrase translation from the secondround returned web page snippets with phonetic, semantic and frequencydistance features. We achieve 46% phrase translation accuracy when using top 10 returned snippets, and 80% accuracy with 165 snippets. Both results are significantly better than several existing methods.
In this paper we describe the components of our statistical machine translation system. This syst... more In this paper we describe the components of our statistical machine translation system. This system combines phrase-tophrase translations extracted from a bilingual corpus using different alignment approaches. Special methods to extract and align named entities are used. We show how a manual lexicon can be incorporated into the statistical system in an optimized way. Experiments on Chinese-to-English and Arabic-to-English translation tasks are presented.
Energy-Based Models (EBMs) capture dependencies between variables by associating a scalar energy ... more Energy-Based Models (EBMs) capture dependencies between variables by associating a scalar energy to each configuration of the variables. Inference consists in clamping the value of observed variables and finding configurations of the remaining variables that minimize the energy. Learning consists in finding an energy function in which observed configurations of the variables are given lower energies than unobserved ones. The EBM approach provides a common theoretical framework for many learning models, including traditional discriminative and generative approaches, as well as graph-transformer networks, conditional random fields, maximum margin Markov networks, and several manifold learning methods.
Probabilistic graphical models associate a probability to each configuration of the relevant vari... more Probabilistic graphical models associate a probability to each configuration of the relevant variables. Energy-based models (EBM) associate an energy to those configurations, eliminating the need for proper normalization of probability distributions. Making a decision (an inference) with an EBM consists in comparing the energies associated with various configurations of the variable to be predicted, and choosing the one with the smallest energy. Such systems must be trained discriminatively to associate low energies to the desired configurations and higher energies to undesired configurations. A wide variety of loss function can be used for this purpose. We give sufficient conditions that a loss function should satisfy so that its minimization will cause the system to approach to desired behavior. We give many specific examples of suitable loss functions, and show an application to object recognition in images. it is important to note that the energy is quantity minimized
International Conference on Multimedia Computing and Systems/International Conference on Multimedia and Expo, 2000
Describes a real-time face-tracking algorithm. We start with single face tracking based on statis... more Describes a real-time face-tracking algorithm. We start with single face tracking based on statistical color modeling and a deformable template. We then expand the algorithm to track multiple faces, possibly with occlusion, by constraining the speed and size changes of the faces. We test the algorithm on sequences with different occlusion patterns and analyze the tracking performance. We also present a tracking software library based on this algorithm. This library can be applied to human-computer interfaces, lip-reading and virtual environments
We present an unsupervised method for learning a hierarchy of sparse feature detectors that are i... more We present an unsupervised method for learning a hierarchy of sparse feature detectors that are invariant to small shifts and distortions. The resulting feature extractor consists of multiple convolution filters, followed by a feature-pooling layer that computes the max of each filter output within adjacent windows, and a point-wise sigmoid non-linearity. A second level of larger and more invariant features is obtained by training the same algorithm on patches of features from the first level. Training a supervised classifier on these features yields 0.64% error on MNIST, and 54% average recognition rate on Caltech 101 with 30 training samples per category. While the resulting architecture is similar to convolutional networks, the layer-wise unsupervised training procedure alleviates the over-parameterization problems that plague purely supervised learning procedures, and yields good performance with very few labeled training samples.
We assess the applicability of several popular learning methods for the problem of recognizing ge... more We assess the applicability of several popular learning methods for the problem of recognizing generic visual categories with invariance to pose, lighting, and surrounding clutter. A large dataset comprising stereo image pairs of 50 uniform-colored toys under 36 azimuths, 9 elevations, and 6 lighting conditions was collected (for a total of 194,400 individual images). The objects were 10 instances of 5 generic categories: four-legged animals, human figures, airplanes, trucks, and cars. Five instances of each category were used for training, and the other five for testing. Low-resolution grayscale images of the objects with various amounts of variability and surrounding clutter were used for training and testing. Nearest neighbor methods, support vector machines, and convolutional networks, operating on raw pixels or on PCA-derived features were tested. Test error rates for unseen object instances placed on uniform backgrounds were around 13% for SVM and 7% for convolutional nets. On a segmentation/recognition task with highly cluttered images, SVM proved impractical, while convolutional nets yielded 16/7% error. A real-time version of the system was implemented that can detect and classify objects in natural scenes at around 10 frames per second.
The detection and recognition of generic object categories with invariance to viewpoint, illumina... more The detection and recognition of generic object categories with invariance to viewpoint, illumination, and clutter requires the combination of a feature extractor and a classifier. We show that architectures such as convolutional networks are good at learning invariant features, but not always optimal for classification, while Support Vector Machines are good at producing decision surfaces from wellbehaved feature vectors, but cannot learn complicated invariances. We present a hybrid system where a convolutional network is trained to detect and recognize generic objects, and a Gaussian-kernel SVM is trained from the features learned by the convolutional network.
Uploads
Papers by Fei HUANG