公开(公告)号：US20220130414A1

公开(公告)日：2022-04-28

申请号：US17504556

申请日：2021-10-19

Applicant: Tata Consultancy Services Limited

Inventor： Ramesh Kumar RAMAKRISHNAN ,  Venkata Subramanian VIRARAGHAVAN ,  Rahul Dasharath GAVAS ,  Sachin PATEL ,  Gauri DESHPANDE

IPC: G10L25/63 ,  G10L25/27 ,  G10L15/04 ,  G06N20/00

Abstract: An important task in several wellness applications is detection of emotional valence from speech. Two types of features of speech signals are used to detect valence: acoustic features and text features. Acoustic features are derived from short frames of speech, while text features are derived from the text transcription. Present disclosure provides systems and methods that determine the effect of text on acoustic features. Acoustic features of speech segments carrying emotion words are to be treated differently from other segments that do not carry such words. Only specific speech segments of the input speech signal are considered based on a dictionary specific to a language to assess emotional valence. A model trained (or trained classifier) for specific language either by including the acoustic features of the emotion related words or by omitting it is used by the system for determining emotional valence in an input speech signal.

公开(公告)号：US20210216909A1

公开(公告)日：2021-07-15

申请号：US17036254

申请日：2020-09-29

Applicant: Tata Consultancy Services Limited

Inventor： Deepa ADIGA ,  Maitry BHAVSAR ,  Mayuri DUGGIRALA ,  Sachin PATEL

IPC: G06N20/00 ,  G06F40/284 ,  G06F40/253 ,  G06F40/268 ,  G06K9/62

Abstract: This disclosure relates generally to methods and systems for automatic extraction of self-reported activities of an individual from a freestyle narrative text. Manual extraction of such self-reported activities of the individual from the freestyle narrative text over the period of time is a complex task and consume a significant amount of time. The present systems and methods utilize a predefined grammar pattern and a natural language processing technique to generate one or more candidate activity phrases, from the pre-processed input text posted by the individual. A deep learning based supervised classification model is utilized to automatically extract the one or more self-reported activities of the individual, from the one or more candidate activity phrases. Manual intervention and efforts of analyzing the freestyle narrative text to extract the self-reported activities are avoided. Longitudinal assessment of the self-reported activities may reveal routines and behavior of the individual.

公开(公告)号：US20200286506A1

公开(公告)日：2020-09-10

申请号：US16812757

申请日：2020-03-09

Applicant: Tata Consultancy Services Limited

Inventor： Gauri Ashutosh DESHPANDE ,  Sachin PATEL ,  Mayuri DUGGIRALA ,  Venkata Subramanian VIRARAGHAVAN

IPC: G10L25/63 ,  G10L15/22 ,  G10L15/02 ,  G10L15/06

Abstract: This disclosure relates generally to speech signal processing, and more particularly to a method and system for processing speech signal for emotion identification. The system processes a speech signal collected as input, during which a plurality of differential features corresponding to a plurality of frames of the speech signal are extracted. Further, the differential features are compared with an emotion recognition model to identify at least one emotion matching the speech signal, and then the at least one emotion is associated with the speech signal.