US 2025/0060284 A1Published application

System and method for transform based subspace interpolation for unsupervised domain adaptation for machine inspection

AssigneeTata Consultancy Services Limited

InventorsKriti KUMAR, Mariswamy Girish CHANDRA, Achanna Anil KUMAR, Angshul MAJUMDAR

Life of this patent

Estimating the term…

Loading patent…

Abstract

This disclosure provides a system and method for transform based subspace interpolation for unsupervised domain adaptation for machine inspection. Embodiments of the present disclosure present a deep transform-based subspace interpolation method to cater to challenging unsupervised adaptation scenario for machine inspection of different but related machines. In the present disclosure, source and target domain data are modeled as low-dimensional subspace using deep transforms. The intermediate domains connecting the two domains are then learned to generate domain invariant features for cross-domain classification. The requisite formulation employing deep transform learning and the closed-form updates for the transforms and their corresponding coefficients are presented. The method of the present disclosure demonstrates potential in learning reliable data representations, particularly in limited data scenario and real-life industrial applications requiring adaptation between different machines.

Claims13

Dependent claims

A processor implemented method, comprising:

obtaining, via one or more hardware processors, (i) a plurality of source domain data (Xs) comprising a first set of features (d) with a plurality of measurements (ns), and (ii) a plurality of target domain data (Xt) comprising a second set of features (d) with a plurality of measurements (nt) for a machine inspection;
learning, via the one or more hardware processors, a deep source domain transform (T0) associated with the plurality of source domain data using a N-layer deep transform learning (DTL) architecture, wherein a transform for each layer of N-layer deep source domain transform (T0) is computed for learning the deep source domain transform (T0);
computing, via the one or more hardware processors, a plurality of coefficients (Z0) on the plurality of source domain data (Xs) using the deep source domain transform (T0);
and
learning, via one or more hardware processors, a set of intermediate deep transforms associated with a plurality of intermediate domains, wherein the set of intermediate deep transforms associated with the plurality of intermediate domains are learned by iteratively transforming the plurality of target domain data (Xt) along a direction that reduces a residue on the plurality of target domain data (Xt) till a deep target domain transform (TM) is obtained that best represents the plurality of target domain data (Xt), and wherein step of iteratively transforming the plurality of target domain data (Xt) comprises:
(i) computing, a plurality of coefficients (Zm) corresponding to the plurality of target domain data (Xt) by transforming the plurality of target domain data (Xt) using a transform of current subspace;
(ii) computing the residue on the plurality of target domain data (Xt) using a current transform and the plurality of coefficients (Zm) corresponding to the plurality of target domain data (Xt);
(iii) computing a change in transform of a current subspace to reduce the residue on the plurality of target domain data (Xt);
and
(iv) computing a transform of a subsequent subspace by adding the change in the transform to the transform of the current subspace.

7 more elements

A system, comprising:

a memory storing instructions;
one or more communication interfaces;
and
one or more hardware processors coupled to the memory via the one or more communication interfaces, wherein the one or more hardware processors are configured by the instructions to:
obtain (i) a plurality of source domain data (Xs) comprising a first set of features (d) with a plurality of measurements (ns), and (ii) a plurality of target domain data (Xt) comprising a second set of features (d) with a plurality of measurements (nt) for a machine inspection;
learn a deep source domain transform (T0) associated with the plurality of source domain data using a N-layer deep transform learning (DTL) architecture, wherein a transform for each layer of N-layer deep source domain transform (T0) is computed for learning the deep source domain transform (T0);
compute a plurality of coefficients (Z0) on the plurality of source domain data (Xs) using the deep source domain transform (T0);
and
learn a set of intermediate deep transforms associated with a plurality of intermediate domains, wherein the set of intermediate deep transforms associated with the plurality of intermediate domains are learned by iteratively transforming the plurality of target domain data (Xt) along a direction that reduces a residue on the plurality of target domain data (Xt) till a deep target domain transform (TM) is obtained that best represents the plurality of target domain data (Xt), and wherein step of iteratively transforming the plurality of target domain data (Xt) comprises:
(i) computing, a plurality of coefficients (Zm) corresponding to the plurality of target domain data (Xt) by transforming the plurality of target domain data (Xt) using a transform of current subspace;
(ii) computing the residue on the plurality of target domain data (Xt) using a current transform and the plurality of coefficients (Zm) corresponding to the plurality of target domain data (Xt);
(iii) computing a change in transform of a current subspace to reduce the residue on the plurality of target domain data (Xt);
and
(iv) computing a transform of a subsequent subspace by adding the change in the transform to the transform of the current subspace.

11 more elements

One or more non-transitory computer readable mediums comprising one or more instructions which when executed by one or more hardware processors cause:

obtaining (i) a plurality of source domain data (Xs) comprising a first set of features (d) with a plurality of measurements (ns), and (ii) a plurality of target domain data (Xt) comprising a second set of features (d) with a plurality of measurements (nt) for a machine inspection;
learning a deep source domain transform (T0) associated with the plurality of source domain data using a N-layer deep transform learning (DTL) architecture, wherein a transform for each layer of N-layer deep source domain transform (T0) is computed for learning the deep source domain transform (T0);
computing a plurality of coefficients (Z0) on the plurality of source domain data (Xs) using the deep source domain transform (T0);
and
learning a set of intermediate deep transforms associated with a plurality of intermediate domains, wherein the set of intermediate deep transforms associated with the plurality of intermediate domains are learned by iteratively transforming the plurality of target domain data (Xt) along a direction that reduces a residue on the plurality of target domain data (Xt) till a deep target domain transform (TM) is obtained that best represents the plurality of target domain data (Xt), and wherein step of iteratively transforming the plurality of target domain data (Xt) comprises:
(i) computing, a plurality of coefficients (Zm) corresponding to the plurality of target domain data (Xt) by transforming the plurality of target domain data (Xt) using a transform of current subspace;
(ii) computing the residue on the plurality of target domain data (Xt) using a current transform and the plurality of coefficients (Zm) corresponding to the plurality of target domain data (Xt);
(iii) computing a change in transform of a current subspace to reduce the residue on the plurality of target domain data (Xt);
and
(iv) computing a transform of a subsequent subspace by adding the change in the transform to the transform of the current subspace.

7 more elements

Dependencies are inferred from the claim text.