Hash codes for images

Patent metadata

Field	Value
Publication number	US-2016358043-A1
Application number	US-201514732463-A
Country	US
Kind code	A1
Filing date	Jun 5, 2015
Priority date	Jun 5, 2015
Publication date	Dec 8, 2016
Grant date	—

Abstract

A method includes receiving, at a neural network, a subset of images of a plurality of images of a training image set. The method includes training the neural network by iteratively adjusting parameters of the neural network based on concurrent application of multiple loss functions to the subset of images. The multiple loss functions include a classification loss function and a hashing loss function. The classification loss function is associated with an image classification function that extracts image features from an image. The hashing loss function is associated with a hashing function that generates a hash code for the image.

First claim

What is claimed is:

1 . A method comprising:
receiving, at a neural network, a subset of images of a plurality of images of a training image set; and training the neural network by iteratively adjusting parameters of the neural network based on concurrent application of multiple loss functions to the subset of images, the multiple loss functions including a classification loss function and a hashing loss function, wherein the classification loss function is associated with an image classification function that extracts image features from an image, and wherein the hashing loss function is associated with a hashing function that generates a hash code for the image.

2 . The method of claim 1 , further comprising generating, based on the adjusted parameters of the neural network, a deep hash model that includes a plurality of deep hash model parameters.

3 . The method of claim 2 , further comprising sending the plurality of deep hash model parameters to a mobile device, the mobile device to execute an image hashing application based on the plurality of deep hash model parameters to generate an image signature for an image file stored at the mobile device, wherein the image signature has a first number of bits that is less than a second number of bits associated with the image file.

4 . The method of claim 2 , further comprising sending the plurality of deep hash model parameters to an image service server, the image service server to utilize the plurality of deep hash model parameters to perform an image service operation.

5 . The method of claim 4 , wherein the image service server is configured to send a set of images to a mobile device, the set of images retrieved from an image search database based on a comparison of an image signature received from the mobile device to a plurality of image signatures generated at the image service server based on the plurality of deep hash model parameters.

6 . The method of claim 5 , wherein the set of images includes a set of Joint Photographic Experts Group (JPEG) images.

7 . The method of claim 6 , wherein the image search database includes an index that is generated based on the plurality of deep hash model parameters, and wherein the index is utilized to identify the set of images based on the image signature.

8 . The method of claim 1 , wherein, during a training stage, the parameters of the neural network are iteratively adjusted based on concurrent application of the multiple loss functions to each image of the plurality of images of the training image set.

9 . The method of claim 8 , wherein a number of parameter adjustment iterations for an individual image of the plurality of images is determined based on a parameter adjustment iteration threshold.

10 . The method of claim 8 , wherein, during the training stage, a first task and a second task are performed concurrently in top layers of the neural network, wherein the first task is associated with a hashing loss layer corresponding to the hashing function and the hashing loss function, and wherein the second task is associated with an image classification layer corresponding to the image classification function and the classification loss function.

11 . The method of claim 10 , wherein the first task includes adjusting a first set of parameters associated with the hashing function based on a first loss value generated by the hashing loss function, and wherein the second task includes adjusting a second set of parameters associated with the image classification function based on a second loss value generated by the classification loss function.

12 . The method of claim 10 , further comprising, after the training stage, generating a deep hash model that includes the first set of parameters associated with the hashing function.

13 . The method of claim 12 , wherein a subset of parameters of the second set of parameters associated with the image classification function are not utilized for generating hash codes during the training stage, and wherein the deep hash model does not include the subset of parameters.

14 . A system comprising:
a processor; and a memory coupled to the processor, the memory storing instructions that, when executed by the processor, cause the processor to perform operations including:
during a training stage, concurrently perform a first task and a second task at a neural network for multiple subsets of images of a plurality of images of a training image set, wherein the first task includes adjusting a first set of parameters associated with a hashing function based on a first loss value generated by a hashing loss function for a particular image, and wherein the second task includes adjusting a second set of parameters associated with an image classification function based on a second loss value generated by a classification loss function for the particular image; and
after the training stage, generating a deep hash model that includes a plurality of deep hash model parameters, wherein the plurality of deep hash model parameters correspond to the adjusted first set of parameters associated with the hashing function.

15 . The system of claim 14 , wherein the operations further comprise sending the plurality of deep hash model parameters to a mobile device, the mobile device to execute an image hashing application based on the plurality of deep hash model parameters to generate an image signature for an image file stored at the mobile device, wherein the image signature has a first number of bits that is less than a second number of bits associated with the image file.

16 . The system of claim 14 , wherein the operations further comprise sending the plurality of deep hash model parameters to an image service server, the image service server to utilize the plurality of deep hash model parameters to perform an image search operation.

17 . The system of claim 14 , wherein the neural network comprises:
convolution layers; pooling layers coupled to the convolution layers; inner product layers coupled to the pooling layers; and top layers coupled to the inner product layers, the top layers to concurrently perform the first task and the second task for the multiple subsets of images.

18 . A mobile device comprising:
a processor; a memory coupled to the processor, the memory storing instructions that, when executed by the processor, cause the processor to perform operations including:
receiving a plurality of deep hash model parameters of a deep hash model; and
utilizing an image hashing application to generate, based on the plurality of deep hash model parameters, an image signature for an image file, the image signature having a first number of bits that is less than a second number of bits associated with the image file,
wherein the plurality of deep hash model parameters are obtained, during a training stage of a neural network, by iteratively adjusting parameters of the neural network based on concurrent application of multiple loss functions to a subset of images of a plurality of images of a training image set, the multiple loss functions including a classification loss function and a hashing loss function, wherein the classification loss function is associated with an image classification function that extracts image features from an image, and wherein the hashing loss function is associated with a hashing function that generates a hash code for the image.

19 . The mobile device of claim 18 , wherein the operations further includes sending the image signature to an image search

Assignees

• At & T Ip I Lp

Inventors

• Mu Yadong
• Liu Zhu

Classifications

• G06V10/70

  using pattern recognition or machine learning (optical pattern recognition or electronic computations therefor G06V10/88) · CPC title

• G06V30/194

  References adjustable by an adaptive method, e.g. learning · CPC title

• G06F16/51 Primary

  Indexing; Data structures therefor; Storage structures · CPC title

• G06N3/045

  Combinations of networks · CPC title

• G06F18/21

  Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation · CPC title