18047562. MEDIA SEGMENT REPRESENTATION USING FIXED WEIGHTS simplified abstract (QUALCOMM Incorporated)
Contents
- 1 MEDIA SEGMENT REPRESENTATION USING FIXED WEIGHTS
- 1.1 Organization Name
- 1.2 Inventor(s)
- 1.3 MEDIA SEGMENT REPRESENTATION USING FIXED WEIGHTS - A simplified explanation of the abstract
- 1.4 Simplified Explanation
- 1.5 Potential Applications
- 1.6 Problems Solved
- 1.7 Benefits
- 1.8 Potential Commercial Applications
- 1.9 Possible Prior Art
- 1.10 Unanswered Questions
- 1.11 Original Abstract Submitted
MEDIA SEGMENT REPRESENTATION USING FIXED WEIGHTS
Organization Name
Inventor(s)
Stephane Villette of San Diego CA (US)
Daniel Jared Sinder of San Diego CA (US)
MEDIA SEGMENT REPRESENTATION USING FIXED WEIGHTS - A simplified explanation of the abstract
This abstract first appeared for US patent application 18047562 titled 'MEDIA SEGMENT REPRESENTATION USING FIXED WEIGHTS
Simplified Explanation
The abstract of the patent application describes a device that stores sets of weights representing media segments and uses processors to generate estimates of associated media segments based on input speech segments.
- The device includes a memory to store sets of weights representing media segments.
- Processors generate data representing input speech segments and pass it into memory units containing sets of weights.
- The processors also generate estimates of associated media segments corresponding to the input speech segments.
Potential Applications
This technology could be applied in speech recognition systems, content recommendation algorithms, and personalized advertising platforms.
Problems Solved
This technology helps improve the accuracy of speech-to-text conversion, enhances content personalization, and optimizes targeted advertising.
Benefits
The benefits of this technology include more accurate speech recognition, personalized content recommendations, and more effective targeted advertising campaigns.
Potential Commercial Applications
Potential commercial applications of this technology include speech recognition software, personalized content platforms, and targeted advertising solutions.
Possible Prior Art
One possible prior art for this technology could be machine learning algorithms used in speech recognition systems and content recommendation engines.
Unanswered Questions
How does this technology handle variations in speech patterns or accents?
The patent abstract does not provide information on how the device adapts to different speech patterns or accents.
What is the computational complexity of the processors in generating estimates of associated media segments?
The abstract does not specify the computational resources required for the processors to generate estimates of associated media segments.
Original Abstract Submitted
A device includes a memory configured to store a collection of sets of weights, each of the sets of weights representing a respective media segment. The device also includes one or more processors configured to generate data representing the detected first input speech segment and to pass the data representing the detected first input speech segment into a collection of memory units. Each memory unit of the collection of memory units includes a set of weights from the collection of sets of weights. The one or more processors are also configured to generate a first estimate of an associated media segment that represents the detected first input speech segment. The associated media segment corresponds to a first memory unit in the collection of memory units.