Showing papers by "John Platt published in 2002"

Auto playlist generation with multiple seed songs

Abstract: The present invention relates to systems and/or methods that generate playlist(s) for a library or collection of media items via selecting a plurality of seed items, at least one of which is an undesirable seed item. Some of the seed items are desirable indicating that a user prefers additional media items similar to the desirable seed items and others are undesirable indicating that the user prefers additional media items dissimilar to the undesirable seed items. Additionally, the seed items can be weighted to establish a relative importance of the seed items. The invention compares media items in the collection with the seed items and determines which media items are added into the playlist by computation of similarity metrics or values. The playlist can be regenerated by adding desirable seed items to the playlist and removing media items from the playlist (e.g., undesirable seed items).

Extracting noise-robust features from audio data

Abstract: A key problem faced by audio identification, classification, and retrieval systems is the mapping of high-dimensional audio input data into informative lower-dimensional feature vectors. This paper explores an automatic dimensionality reduction algorithm called Distortion Discriminant Analysis (DDA). Each layer of DDA projects its input into directions which maximize the SNR for a given set of distortions. Multiple layers efficiently extract features over a wide temporal window. The audio input to DDA undergoes perceptually-relevant preprocessing and de-equalization, to further suppress distortions. We apply DDA to the task of identifying audio clips in an incoming audio stream, based on matching stored audio fingerprints. We show excellent test results on matching input fingerprints against 36 hours of stored audio data.

Probability estimate for K-nearest neighbor

Abstract: Systems and methods are disclosed that facilitate producing probabilistic outputs also referred to as posterior probabilities. The probabilistic outputs include an estimate of classification strength. The present invention intercepts non-probabilistic classifier output and applies a set of kernel models based on a softmax function to derive the desired probabilistic outputs. Such probabilistic outputs can be employed with handwriting recognition where the probability of a handwriting sample classification is combined with language models to make better classification decisions.

System and method for noise-robust feature extraction

Abstract: Extracting features from signals for use in classification, retrieval, or identification of data represented by those signals uses a “Distortion Discriminant Analysis” (DDA) of a set of training signals to define parameters of a signal feature extractor. The signal feature extractor takes signals having one or more dimensions with a temporal or spatial structure, applies an oriented principal component analysis (OPCA) to limited regions of the signal, aggregates the output of multiple OPCAs that are spatially or temporally adjacent, and applies OPCA to the aggregate. The steps of aggregating adjacent OPCA outputs and applying OPCA to the aggregated values are performed one or more times for extracting low-dimensional noise-robust features from signals, including audio signals, images, video data, or any other time or frequency domain signal. Such extracted features are useful for many tasks, including automatic authentication or identification of particular signals, or particular elements within such signals.

Type size dependent anti-aliasing in sub-pixel precision rendering systems

Abstract: Methods and systems for selecting different anti-aliasing filters to be applied to objects of image data rendered with pixel sub-component precision based on the attributes of the objects of the graphics image. For example, when rendering a character traditional anti-aliasing applied in the direction parallel to the striping of the pixel sub-components is adapted to the unique requirements of different character font sizes such that a large amount of anti-aliasing is applied to large font sizes while a small amount of anti-aliasing is provided for characters at small font sizes. Displaced sampling is utilized to determine the alpha values for a plurality of vertical samples for each RGB pixel sub-component. The selected anti-aliasing filter is applied to the alpha values to determine the luminous intensity values of the RGB pixel sub-components.

QuickStroke: an incremental on-line Chinese handwriting recognition system

Abstract: This paper presents QuickStroke: a system for the incremental recognition of handwritten Chinese characters. Only a few strokes of an ideogram need to be entered in order for a character to be successfully recognized. Incremental recognition is a new approach for on-line recognition of ideographic characters. It allows a user to enter characters a factor of 2 times faster than systems that require entry of full characters. Incremental recognition is performed by a two-stage system which utilizes 68 neural networks with more than 5 million free parameters. To enable incremental recognition, we use specialized time-delay neural networks (TDNNs) that are trained to recognize partial characters. To boost the recognition accuracy of complete characters, we also use standard fully-connected neural networks. Quickstroke is 97.3% accurate for the incremental writer-independent recognition of 4400 simplified GB Chinese ideograms.