Huawei technologies co., ltd. (20240126808). SEARCH RESULT FEEDBACK METHOD AND APPARATUS, AND STORAGE MEDIUM simplified abstract
Contents
SEARCH RESULT FEEDBACK METHOD AND APPARATUS, AND STORAGE MEDIUM
Organization Name
Inventor(s)
SEARCH RESULT FEEDBACK METHOD AND APPARATUS, AND STORAGE MEDIUM - A simplified explanation of the abstract
This abstract first appeared for US patent application 20240126808 titled 'SEARCH RESULT FEEDBACK METHOD AND APPARATUS, AND STORAGE MEDIUM
Simplified Explanation
The patent application describes a method for determining the arrangement orders of objects in an image based on scene intent weight, confidence score, and object relationship score.
- The method involves obtaining an image with multiple objects, determining the arrangement orders of the objects based on various scores, and providing search results according to these orders.
- Scene intent weight indicates the probability of an object being searched in a corresponding scene, confidence score measures similarity to images in a library, and object relationship score indicates the importance of an object in the image.
- Potential Applications:
- Image search and retrieval systems - Object recognition in computer vision applications
- Problems Solved:
- Efficient organization of search results - Improved accuracy in object recognition
- Benefits:
- Enhanced search capabilities - Better user experience in image retrieval
- Potential Commercial Applications:
- E-commerce platforms for product search - Security systems for object detection
- Possible Prior Art:
- Existing image recognition algorithms - Similar methods for object arrangement in images
- Unanswered Questions:
- How does the method handle complex scenes with multiple objects of varying sizes and orientations?
- Unanswered Questions:
The patent application does not provide specific details on how the method adapts to complex scenes with diverse objects.
- What is the computational complexity of the algorithm used in determining arrangement orders based on the scene intent weight, confidence score, and object relationship score?
The patent application does not discuss the computational resources required for implementing the method and the efficiency of the algorithm in real-time applications.
Original Abstract Submitted
in one example method, a first image including m objects is obtained. for n objects in the m objects, when n is greater than or equal to 2, arrangement orders of the n objects is determined, where an arrangement order of any one of the n objects is determined based on at least one of a scene intent weight, a confidence score, or an object relationship score. the scene intent weight is used to indicate a probability that the any object is searched in a scene corresponding to the first image, the confidence score is a similarity between the any object and an image in an image library, and the object relationship score is used to indicate importance of the any object in the first image. search results of some or all of the n objects are fed back according to the arrangement orders of the n objects.
