Method and system for name-face/voice-role association
View Patent ↗A method for providing name-face/voice-role association includes determining whether a closed captioned text accompanies a video sequence, providing one of text recognition and speech to text conversion to the video sequence to generate a role-name versus actor-name list from the video sequence, extracting face boxes from the video sequence and generating face models, searching a predetermined portion of text for an entry on the role-name versus actor-name list, searching video frames for face models/voice models that correspond to the text searched by using a time code so that the video frames correspond to portions of the text where role-names are detected, assigning an equal level of certainty for each of the face models found, using lip reading to eliminate face models found that pronounce a role-name corresponding to said entry on the role-name versus actor-name list, scanning a remaining portion of text provided and updating a level of certainty for said each of the face models previously found. Once a particular face model/voice model and role-name association has reached a threshold the role-name, actor name, and particular face model/voice model is stored in a database and can be displayed by a user when the threshold for the particular face model has been reached. Thus the user can query information by entry of role-name, actor name, face model, or even words spoken by the role-name as a basis for the association. A system provides hardware and software to perform these functions.
1. A method for providing name-face/voice-role association, comprising the steps of:
(a) determining whether a closed captioned text accompanies a video sequence;
(b) providing one of text recognition and speech to text conversion to the video sequence to generate a role-name versus actor-name list from the video sequence;
(c) extracting face boxes/voices from the video sequence and generating face models/voice models;
(d) searching a predetermined portion of text provided in step (b) for an entry on the role-name versus actor-name list;
(e) searching video frames for face models/voice models that correspond to the text searched in step (d) by using a time code so that the video frames correspond to portions of the text where role-names are detected;
(f) assigning an equal level of certainty for each of the face models/voice models found in step (e);
(g) using lip reading to eliminate face models found in step (e) that pronounce a role-name corresponding to said entry on the role-name versus actor-name list;
(h) scanning a remaining portion of text provided in step (b) and updating a level of certainty for said each of the face models/voice models found in step (e);
(i) determining whether a particular face model/voice model and role-name association has reached a threshold;
(j) storing the role-name, actor name, and particular face model/voice model in a database when the threshold for the particular face model/voice model has been reached.
2. The method according to claim 1 , further comprising:
(k) repeating steps d through j for each entry on the role-name versus actor-name list.
3. The method according to claim 1 , wherein step (j) includes
(i) backpropagating and marking all video segments of the video sequence containing the particular face model/voice model.
4. The method according to claim 1 , wherein the extracting of face boxes in step (c) is performed using an eigenvector based method for face matching.
5. The method according to claim 1 , wherein the extracting of face boxes is performed by using model-based face extraction.
6. The method according to claim 1 , wherein the voice models are determined by using MFCC (Mel frequency cepstral coefficients).
7. A system for providing name-face-role association, comprising:
a processor;
storage means for the processor;
a database which is accessible by the processor;
means for detecting closed captioned text of a program;
means for extracting face boxes and generating face models/voice models of the program;
a search engine used by the processor for searching the program by role-name versus actor-name for a particular role name;
lip reading detection means for identifying a face model of the particular role-name in the program by eliminating face models which pronounce the particular role name;
communication means for providing a user with the identity of the particular role-name;
means to update the database with the face model/voice model of the particular role-name associated with actor name.
8. The system according to claim 7 , further comprising speech-to-text conversion means for use in the absence of closed-captioned text.
9. The system according to claim 7 , wherein the processor, means for detecting closed captioned text, means for extracting face boxes, and the search engine are arranged in a network server.
10. The system according to claim 9 , wherein the communication means between the user and the system is the Internet.
11. The system according to claim 9 , wherein the communication means between the user and the system is one of fiber optic and RF.
12. The system according to claim 11 , wherein the particular role-name provided to the user by the communication means is communicated to the user in HTML format.
13. The system according to claim 7 , wherein the program in containing the role-name versus actor name is one of broadcast, videotape, videodisc, and videostream.
14. The system according to claim 7 , wherein the system comprises a home video system.
15. The system according to claim 7 , wherein the system comprises a teleconferencing system.