Peter L De

US Patent:

7219056, May 15, 2007

Filed:

Apr 19, 2001

Appl. No.:

09/838449

Inventors:

Scott Elliot Axelrod - Mt. Kisco NY, US
Peder Andreas Olsen - New York NY, US
Harry William Printz - San Francisco CA, US
Peter Vincent de Souza - San Jose CA, US

Assignee:

International Business Machines Corporation - Armonk NY

International Classification:

G10L 15/00
G06F 7/60

US Classification:

704235, 703 2, 704236, 704246, 704251

Abstract:

Two statistics are disclosed for determining the quality of language models. These statistics are called acoustic perplexity and the synthetic acoustic word error rate (SAWER), and they depend upon methods for computing the acoustic confusability of words. It is possible to substitute models of acoustic data in place of real acoustic data in order to determine acoustic confusability. An evaluation model is created, a synthesizer model is created, and a matrix is determined from the evaluation and synthesizer models. Each of the evaluation and synthesizer models is a hidden Markov model. Once the matrix is determined, a confusability calculation may be performed. Different methods are used to determine synthetic likelihoods. The confusability may be normalized and smoothed and methods are disclosed that increase the speed of performing the matrix inversion and the confusability calculation. A method for caching and reusing computations for similar words is disclosed.

US Patent:

8304253, Nov 6, 2012

Filed:

Oct 19, 2006

Appl. No.:

12/090922

Inventors:

Uichong Brandon Yi - Los Angeles CA, US
Peter Patrick De Guzman - Orange CA, US
Wayne Liu - Los Angeles CA, US
Chang-Jin Kim - Beverly Hills CA, US

Assignee:

Advanced Liquid Logic Inc - Research Triangle Park NC

International Classification:

B01D 57/00
B01L 99/00

US Classification:

436177, 436174, 436180, 422500, 422502, 422504

Abstract:

A refill droplet facilitates the extraction of a droplet laterally from a channel in a microfluidic apparatus. Such extraction allows a discrete band of separated particles or solute molecules to be excised from a fluid stream and processed and analyzed separately. An extraction point is located along the length of the channel and includes an EWOD surface or similar microfluidic technology to extract a droplet. An opening in the channel opposite the extraction means is equipped with microfluidic technology to transport a refill droplet to the opening. The refill droplet is moved into the channel or column to occupy the area previously occupied by the extracted droplet. This prevents distortion or mixing of the bands of particles or molecules within the channel and prevents the draining of any portion of the fluidic system.

US Patent:

51691990, Dec 8, 1992

Filed:

Feb 1, 1991

Appl. No.:

7/649418

Inventors:

Peter J. de Ruyter - Northridge CA
John R. Bryant - Van Nuys CA
Roland H. Schneider - Marina del Rey CA

International Classification:

B66C 112

US Classification:

294149

Abstract:

Disclosed is an object carrier comprising a combination of a strap member and a hook-and-fabric-type fastener. The fabric element of the hook-and-fabric-type fastener covers substantially the entire external surface of the strap member. A ring is attached to one end of the strap member and the hook element of the hook-and-fabric-type fastener is attached to the other end. There is a handle on the external surface near the ring. The strap member is wrapped tightly about an object to be carried, with the hook element drawn through the ring and then passed into engagement with the fabric element.

US Patent:

D2611376, Oct 6, 1981

Filed:

Dec 15, 1978

Appl. No.:

5/969875

Inventors:

Peter G. De Lillo - Staten Island NY

International Classification:

D1401

US Classification:

D14 33

US Patent:

50724525, Dec 10, 1991

Filed:

Nov 2, 1989

Appl. No.:

7/431720

Inventors:

Peter F. Brown - New York NY
Peter V. De Souza - Yorktown Heights NY
Michael A. Picheny - White Plains NY

Assignee:

International Business Machines Corporation - Armonk NY

International Classification:

G10L 506

US Classification:

381 43

Abstract:

In a Markov model speech recognition system, an acoustic processor generates one label after another selected from an alphabet of labels. Each vocabulary word is represented as a baseform constructed of a sequence of Markov models. Each Markov model is stored in a computer memory as (a) a plurality of states; (b) a plurality of arcs, each extending from a state to a state with a respective stored probability; and (c) stored label output probabilities, each indicating the likelihood of a given label being produced at a certain arc. Word likelihood based on acoustic characteristics is determined by matching a string of labels generated by the acoustic processor against the probabilities stored for each word baseform. Improved models of words are obtained by specifying label parameters and constructing word baseforms interdependently and iteratively.