Patents
Literature
Patsnap Copilot is an intelligent assistant for R&D personnel, combined with Patent DNA, to facilitate innovative research.
Patsnap Copilot

133 results about "Harmonization" patented technology

In music, harmonization is the chordal accompaniment to a line or melody: "Using chords and melodies together, making harmony by stacking scale tones as triads". A harmonized scale can be created by using each note of a musical scale as a root note for a chord and then by taking other tones within the scale building the rest of a chord.

Systems and methods for recognizing objects in radar imagery

ActiveUS20160019458A1Low in size and weight and power requirementImprove historical speed and accuracy performance limitationDigital computer detailsDigital dataPattern recognitionGraphics
The present invention is directed to systems and methods for detecting objects in a radar image stream. Embodiments of the invention can receive a data stream from radar sensors and use a deep neural network to convert the received data stream into a set of semantic labels, where each semantic label corresponds to an object in the radar data stream that the deep neural network has identified. Processing units running the deep neural network may be collocated onboard an airborne vehicle along with the radar sensor(s). The processing units can be configured with powerful, high-speed graphics processing units or field-programmable gate arrays that are low in size, weight, and power requirements. Embodiments of the invention are also directed to providing innovative advances to object recognition training systems that utilize a detector and an object recognition cascade to analyze radar image streams in real time. The object recognition cascade can comprise at least one recognizer that receives a non-background stream of image patches from a detector and automatically assigns one or more semantic labels to each non-background image patch. In some embodiments, a separate recognizer for the background analysis of patches may also be incorporated. There may be multiple detectors and multiple recognizers, depending on the design of the cascade. Embodiments of the invention also include novel methods to tailor deep neural network algorithms to successfully process radar imagery, utilizing techniques such as normalization, sampling, data augmentation, foveation, cascade architectures, and label harmonization.
Owner:GENERAL DYNAMICS MISSION SYST INC

Under-actuated coupling transmission type imitation human finger mechanism

The invention provides an under-actuated coupling driving finger-imitating mechanism, in particular to a finger-imitating mechanism on the manipulator of a robot. The invention solves the problems of poor harmonization of all knuckles, being difficult to keep holding shape, slow speed when wrapping the article to be held, poor adaptability, etc. when the existing manipulator finger carries out a holding motion. One end of the invention, which is close to a knuckle connecting rod (23), is connected with a basic joint moment sensor (22); the other end close to the knuckle connecting rod (23) is connected with the shaft disc (54) of a central joint shaft; one end of a coupling connecting rod (28) is connected with a left plate (25) close to the knuckle and a right plate (4) close to the knuckle; the other end of the coupling connecting rod (28) is connected with a fingertip (14); a central knuckle left plate (26) and a central knuckle right plate (2) are connected with the shaft disc (54) of the central knuckle shaft by a coupling central knuckle shaft (30). By adopting the coupling transmission way, the central knuckle and the fingertip of the invention have the advantages of good motion harmony when the finger is bent, fast speed and convenient pre-tightening when the article to be held is wrapped.
Owner:HARBIN INST OF TECH

Human-machine voice interaction method and human-machine voice interaction system

The invention discloses a human-machine voice interaction method and a human-machine voice interaction system. The method includes: acquiring voice information of a user, analyzing the voice information to obtain a language type, adopting a recognition mode corresponding to the language type to recognize the voice information, searching for matching the voice information with pre-stored feedback voice information to obtain matched feedback voice information, and searching for matching the recognized voice information with pre-stored feedback video information to obtain matched feedback video information; if the two types of information are in association, synchronously outputting the two types of information. By means of the human-machine voice interaction method, voice of various languages can be recognized to realize interaction of various languages, and video and audio synchronous reaction can be realized as well to achieve better customer experience; further, by judgment of the association between the feedback voice information and the feedback video information, consistency of the feedback voice information and the feedback video information is realized, interaction accuracy is improved, and higher harmonization and synchronization in voice and image can be achieved.
Owner:SHANGHAI FINEKITE EXHIBITION ENG

Systems and methods for recognizing objects in radar imagery

ActiveUS9978013B2Low in size and weight and power requirementImprove historical speed and accuracy performance limitationNeural architecturesRadio wave reradiation/reflectionGraphicsData stream
The present invention is directed to systems and methods for detecting objects in a radar image stream. Embodiments of the invention can receive a data stream from radar sensors and use a deep neural network to convert the received data stream into a set of semantic labels, where each semantic label corresponds to an object in the radar data stream that the deep neural network has identified. Processing units running the deep neural network may be collocated onboard an airborne vehicle along with the radar sensor(s). The processing units can be configured with powerful, high-speed graphics processing units or field-programmable gate arrays that are low in size, weight, and power requirements. Embodiments of the invention are also directed to providing innovative advances to object recognition training systems that utilize a detector and an object recognition cascade to analyze radar image streams in real time. The object recognition cascade can comprise at least one recognizer that receives a non-background stream of image patches from a detector and automatically assigns one or more semantic labels to each non-background image patch. In some embodiments, a separate recognizer for the background analysis of patches may also be incorporated. There may be multiple detectors and multiple recognizers, depending on the design of the cascade. Embodiments of the invention also include novel methods to tailor deep neural network algorithms to successfully process radar imagery, utilizing techniques such as normalization, sampling, data augmentation, foveation, cascade architectures, and label harmonization.
Owner:GENERAL DYNAMICS MISSION SYST INC
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products