Generating a Multi-Use Vocabulary based on Image Data

• US 20080205770A1
• Filed: 07/31/2007
• Published: 08/28/2008
• Est. Priority Date: 02/26/2007
• Status: Active Grant

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Microsoft Corporation
• Inventors: Ma, Wei-Ying, Xie, Xing, Jia, Menglei
• Granted Patent: US 8,396,331 B2
• Time in Patent Office: Days
• US Class Current: 382/217
• CPC Class Codes: G06V 10/464 using a plurality of salient features, e.g. bag-of-words [BoW] representations

Abstract

Functionality is described for generating a vocabulary from a source dataset of image items or other non-textual items. The vocabulary serves as a tool for retrieving items from a target dataset in response to queries. The vocabulary has at least one characteristic that allows it to be used to retrieve items from multiple different target datasets. A target dataset can have a different size than the source dataset and/or a different type than the source dataset. The enabling characteristic may correspond to a size of the source dataset above a prescribed minimum number of items and/or a size of the vocabulary above a prescribed minimum number of words.

Claims

1. A method for generating a vocabulary for non-textual items, comprising:

providing a source dataset comprising a plurality of items;

identifying features in the source dataset; and

generating a plurality of words associated with the features, to thereby form a vocabulary,

wherein the vocabulary serves as a mechanism for use in retrieving items from plural different target datasets in response to queries made to the plural different target datasets.

2. The method of claim 1, wherein the non-textual items comprise image items.

3. The method of claim 1, wherein the source dataset has at least a predetermined size to ensure that the vocabulary can serve as a mechanism for use in retrieving items from plural different target datasets.

4. The method of claim 3, wherein the predetermined size corresponds to an approximate transition point at which further increases in size do not yield significant increases in performance, relative to increases in size prior to the transition point, wherein the vocabulary is generated using a source dataset having approximately only the predetermined size.

5. The method of claim 1, wherein the vocabulary has at least a predetermined number of words to ensure that the vocabulary can serve as a mechanism for use in retrieving items from plural different target datasets.

6. The method of claim 5, wherein the predetermined number corresponds to an approximate transition point at which further increases in number do not yield significant increases in performance, relative to increases in number prior to the transition point, wherein the vocabulary is generated to have approximately only the predetermined number.

7. The method of claim 1, wherein at least one of the target datasets has a larger size than the source dataset.

8. The method of claim 1, wherein at least one of the target datasets has a different type than the source dataset.

9. One or more computing devices, comprising:

one or more processors; and

memory to store computer-executable instructions that, when executed by the one or more processors, perform the method of claim 1.

10. One or more machine-readable media containing machine-readable instructions for implementing the method of claim 1.

12. The method of claim 1, wherein the non-textual items comprise image items.

11. A method for generating a vocabulary for non-textual items, comprising:

determining at least one characteristic of a vocabulary that ensures that the vocabulary can serve as a mechanism for use in retrieving items from plural different target datasets; and

providing a vocabulary that meets said at least one characteristic.

13. The method of claim 11, wherein said at least one characteristic refers to a minimum size of a source dataset used to generate the vocabulary.

14. The method of claim 13, wherein the minimum size corresponds to an approximate transition point at which further increases in size do not yield significant increases in performance, relative to increases in size prior to the transition point, wherein vocabulary is generated using a source dataset having approximately only the predetermined size.

15. The method of claim 11, wherein said at least one characteristic refers to a minimum number of words in the vocabulary.

16. The method of claim 15, wherein minimum number corresponds to an approximate transition point at which further increases in number do not yield significant increases in performance, relative to increases in number prior to the transition point, wherein the vocabulary is generated to have approximately only the predetermined number.

17. One or more computing devices, comprising:

one or more processors; and

memory to store computer-executable instructions that, when executed by the one or more processors, perform the method of claim 11.

18. One or more machine-readable media containing machine-readable instructions for implementing the method of claim 11.

19. A vocabulary providing module implemented by a data processing device configured to provide a vocabulary from a source dataset of image items, wherein the vocabulary has at least one characteristic that ensures that the vocabulary can serve as a mechanism for use in retrieving image items from plural different target datasets, wherein any target dataset can have a different size than the source dataset, and wherein any target dataset can have a different type than the source dataset.

20. The vocabulary providing module of claim 19, wherein said at least one characteristic refers to:

a size of the source dataset used to generate the vocabulary above a minimum size;

or

a number of words in the vocabulary above a minimum number.