• Search stock it
  • Search content
  • Search author
  • Search topics
Hot:China safe it BOE A it Music as Internet cafes
GEM refers to it
Published on 2017-10-19 12:26:23 Fortune iPhone for the East
The company is located in:Report
                        Heavyweight world's first voiceprint recognition artificial intelligence TV listed Changhong set off interactive revolution Pacific Home Network 10-19
                            Heavyweight world's first voiceprint recognition artificial intelligence TV listed Changhong set off interactive revolution
The Pacific OceanHome Network 10-19 09:51
Since the advent of televisions for more than 80 years, the continuous innovation of display technologies, from black and white to color, from 2D to 3D, from standard definition to ultra-high definition, has resulted in more and more beautiful images. However, it seems increasingly difficult to keep the public on television Pre-flight, which is related to the outbreak of the source of content in the Internet era, rich and personalized content presented through the network. In fact, not because people do not like watching TV, but because most people do not have time to watch TV during the daytime, only the fragmentation time is left. Therefore, moving the small screen is more in line with people's needs, and the interactive form is limited to switching channels with the remote control The traditional television began to become more and more distant from the user.
As early as six years ago, Changhong, which started to make power television voice interaction, once again brought epoch-making artificial intelligence CHiQ TV 5K on October 12. This tasting event held in Chengdu will bring its far-field voice and voiceprint Identification, fuzzy retrieval three core selling show the most vividly, especially for the first time Changhong in the artificial intelligence television industry voiceprint identification technology, but also with the industry competitors to widen the gap between the key technologies.
Voiceprint recognition gives television the ability to recognize family members
Voiceprint identification technology, as its name implies, is to identify the identity of the individual who issued the voice command, and then according to everyone's preferences, watch history, recommend audio and video programs. While other competing voice recognition capabilities, but also to ignore the voice commands issued by the key issue of this, only based on historical browsing and semantic understanding of the program to achieve the recommendation, but this is for home users, but it should not be ignored In the current artificial intelligence television technology, user behavior portrait is listed as one of the basic data, but can not subdivide family members, will greatly reduce the accuracy of user behavior analysis.
At the tasting scene, Chen Qiyu, CHiQ's television product manager, and another female guest played the role of family member. When the commentary "I want to watch a movie" was given, the recommended film source given by CHiQ smart TV did not Same, and the interface also selects different pattern colors according to the voiceprint recognition. The interface and recommendation results of adult men, women and children are respectively corresponding to their personality and viewing habits. When the audience who has never used the television set speaks, , CHiQ recognizes that this is a stranger, so videos are only suggested by default.
Chen Keyu introduced that Changhong applied the international forefront of I-vector technology for voiceprint modeling, at the same time with the cloud computing platform to establish a unique voiceprint database for each family member, to quickly and accurately in the home application environment to achieve family Members of the identification, recognition rate of more than 90%.
It is understood that to achieve such an excellent recognition rate is not easy, Changhong voiceprint recognition project team encountered many difficulties in the development process, early in the project, less voiceprint data is one of the major difficulties facing the model can not be established Effective verification and training. The voiceprint recognition algorithm needs enough training data to be able to accurately learn the differences between different speakers. To this end, each member of the project team holding a remote control on time every day, keep shouting "Changhong white ... I want to see ... I want to see ...", again, twice ... countless times, day after day to collect sound Pattern samples.
With enough data, the design of voiceprint recognition algorithms has become even more important. Prior to the project team's accumulation of technology mainly in the field of image recognition, and no voice-related experience, so the algorithm design is particularly difficult. The core members of the project team Professor Wu Ying body in the United States, regularly twice a week through video conferencing and domestic project team members to discuss, after about three months of hard work, finally completed the algorithm design, server deployment, and the successful delivery .
Fuzzy search and far-field speech make artificial intelligence more intimate
Nowadays, many AI televisions collect voice commands from users with a built-in microphone on the remote control, which greatly reduces the consistency of voice interaction. The CHiQ5K is equipped with an integrated array microphone to achieve far-field voice collection within 5 meters and enables voice on / off , From the turn on the TV to complete the program broadcast, the whole without the remote control, only through voice control can be achieved.
The computer can perform various tasks under precise instructions. The value of artificial intelligence lies in identifying natural dialogues and less precise instructions. The fuzzy search capability of Changhong CHiQ5 brings powerful correlation analysis and error correction capabilities, Even if the user's expression is not accurate enough, he can also present the video resources the user wants.
To fuzzy retrieval function, for example, when users on-demand video, inadvertently "Momo" in the "M" word according to the wrong, labeled as "half" word, this time TV intelligent system can determine their own, and According to the user big data error correction keywords, which will be on-demand programs targeted to "芈 month Biography."
To realize this function, Chen Keyu said: "With such interaction requires technical support synergies, behind the Q5K artificial intelligence television, there are basic data platform, semantic platform, user portrait platform and media gathering platform Support.Television with semantic understanding algorithm, search algorithm, recommendation algorithm, data mining algorithms algorithm system.Also there are equipment control protocol, basic application management protocol, data query protocol and algorithm function protocol.With such technical system support artificial intelligence TV Q5K has a strong interaction, growth, thinking, can be widely used in conjunction with other features of the device. "
Voice recognition, far-field voice, fuzzy search, look simple and refined three characteristics behind the accumulation of Changhong years of research and development achievements accumulated, Changhong a new generation of artificial intelligence television Q5K launch, refresh China in the world of artificial intelligence The new height of technology, the establishment of artificial intelligence television technology, a new benchmark for television, a traditional home entertainment center has brought a revolutionary interactive experience.
                    SortThe earliestrecentMost likes
Posted by 2017-10-19 15:04:53
d
Comment on this topicThe post is gone! How to do?
Author: You will not be publishedlog in |5 seconds to register Author:, Welcome messagedrop out |Post new topic
                Tip: All the information and comments made by users in the community represent only personal opinions. They have nothing to do with the position of this website and do not constitute any investment advice for you. Users should make their own decisions on securities investment and bear the corresponding risks based on their own independent judgment."Post comments self-discipline management commitment"
Heavyweight world's first voiceprint recognition artificial intelligence TV listed Changhong set off interactive revolution The Pacific OceanHome Network 10-19 09:51 Since the advent of televisions for more than 80 years, the continuous innovation of display technologies, from black and white to color, from 2D to 3D, from standard definition to ultra-high definition, has resulted in more and more beautiful images. However, it seems increasingly difficult to keep the public on television Pre-flight, which is related to the outbreak of the source of content in the Internet era, rich and personalized content presented through the network. In fact, not because people do not like watching TV, but because most people do not have time to watch TV during the daytime, only the fragmentation time is left. Therefore, moving the small screen is more in line with people's needs, and the interactive form is limited to switching channels with the remote control The traditional television began to become more and more distant from the user. As early as six years ago, Changhong, which started to make power television voice interaction, once again brought epoch-making artificial intelligence CHiQ TV 5K on October 12. This tasting event held in Chengdu will bring its far-field voice and voiceprint Identification, fuzzy retrieval three core selling show the most vividly, especially for the first time Changhong in the artificial intelligence television industry voiceprint identification technology, but also with the industry competitors to widen the gap between the key technologies. Voiceprint recognition gives television the ability to recognize family members Voiceprint identification technology, as its name implies, is to identify the identity of the individual who issued the voice command, and then according to everyone's preferences, watch history, recommend audio and video programs. While other competing voice recognition capabilities, but also to ignore the voice commands issued by the key issue of this, only based on historical browsing and semantic understanding of the program to achieve the recommendation, but this is for home users, but it should not be ignored In the current artificial intelligence television technology, user behavior portrait is listed as one of the basic data, but can not subdivide family members, will greatly reduce the accuracy of user behavior analysis. At the tasting scene, Chen Qiyu, CHiQ's television product manager, and another female guest played the role of family member. When the commentary "I want to watch a movie" was given, the recommended film source given by CHiQ smart TV did not Same, and the interface also selects different pattern colors according to the voiceprint recognition. The interface and recommendation results of adult men, women and children are respectively corresponding to their personality and viewing habits. When the audience who has never used the television set speaks, , CHiQ recognizes that this is a stranger, so videos are only suggested by default. Chen Keyu introduced that Changhong applied the international forefront of I-vector technology for voiceprint modeling, at the same time with the cloud computing platform to establish a unique voiceprint database for each family member, to quickly and accurately in the home application environment to achieve family Members of the identification, recognition rate of more than 90%. It is understood that to achieve such an excellent recognition rate is not easy, Changhong voiceprint recognition project team encountered many difficulties in the development process, early in the project, less voiceprint data is one of the major difficulties facing the model can not be established Effective verification and training. The voiceprint recognition algorithm needs enough training data to be able to accurately learn the differences between different speakers. To this end, each member of the project team holding a remote control on time every day, keep shouting "Changhong white ... I want to see ... I want to see ...", again, twice ... countless times, day after day to collect sound Pattern samples. With enough data, the design of voiceprint recognition algorithms has become even more important. Prior to the project team's accumulation of technology mainly in the field of image recognition, and no voice-related experience, so the algorithm design is particularly difficult. The core members of the project team Professor Wu Ying body in the United States, regularly twice a week through video conferencing and domestic project team members to discuss, after about three months of hard work, finally completed the algorithm design, server deployment, and the successful delivery . Fuzzy search and far-field speech make artificial intelligence more intimate Nowadays, many AI televisions collect voice commands from users with a built-in microphone on the remote control, which greatly reduces the consistency of voice interaction. The CHiQ5K is equipped with an integrated array microphone to achieve far-field voice collection within 5 meters and enables voice on / off , From the turn on the TV to complete the program broadcast, the whole without the remote control, only through voice control can be achieved. The computer can perform various tasks under precise instructions. The value of artificial intelligence lies in identifying natural dialogues and less precise instructions. The fuzzy search capability of Changhong CHiQ5 brings powerful correlation analysis and error correction capabilities, Even if the user's expression is not accurate enough, he can also present the video resources the user wants. To fuzzy retrieval function, for example, when users on-demand video, inadvertently "Momo" in the "M" word according to the wrong, labeled as "half" word, this time TV intelligent system can determine their own, and According to the user big data error correction keywords, which will be on-demand programs targeted to "芈 month Biography." To realize this function, Chen Keyu said: "With such interaction requires technical support synergies, behind the Q5K artificial intelligence television, there are basic data platform, semantic platform, user portrait platform and media gathering platform Support.Television with semantic understanding algorithm, search algorithm, recommendation algorithm, data mining algorithms algorithm system.Also there are equipment control protocol, basic application management protocol, data query protocol and algorithm function protocol.With such technical system support artificial intelligence TV Q5K has a strong interaction, growth, thinking, can be widely used in conjunction with other features of the device. " Voice recognition, far-field voice, fuzzy search, look simple and refined three characteristics behind the accumulation of Changhong years of research and development achievements accumulated, Changhong a new generation of artificial intelligence television Q5K launch, refresh China in the world of artificial intelligence The new height of technology, the establishment of artificial intelligence television technology, a new benchmark for television, a traditional home entertainment center has brought a revolutionary interactive experience.