Topic segmentation method and device based on deep learning, equipment and medium
A deep learning and topic technology, applied in neural learning methods, character and pattern recognition, instruments, etc., can solve problems such as low accuracy rate and rough topic outline, and achieve the effect of improving efficiency and accuracy, and improving the effect of topic segmentation
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
Embodiment 1
[0064] see figure 1 , figure 1 It is a schematic flowchart of a topic segmentation method disclosed in an embodiment of the present invention. Such as figure 1 As shown, the topic segmentation method includes the following steps:
[0065] 110. Create and train a deep learning-based instance segmentation model.
[0066] The instance segmentation model can be built using a deep learning-based instance segmentation network, for example, the MaskR-CNN instance segmentation network can be used. The Mask R-CNN instance segmentation network adopts a flexible and general target instance segmentation framework, which can detect objects and perform instance segmentation at the same time. The object mask network is added on the basis of the original Faster R-CNN, and the speed is about 5FPS.
[0067] Firstly, an initial model of instance segmentation based on deep learning is created, and the network parameters of the initial model of instance segmentation are randomly initialized. ...
Embodiment 2
[0091] see image 3 , image 3 It is a structural schematic diagram of a topic segmentation device disclosed in an embodiment of the present invention. Such as image 3 As shown, the topic segmentation device may include:
[0092] The creation unit 210 is used to create and train an instance segmentation model based on deep learning;
[0093] A segmentation unit 220, configured to obtain an input image, and segment the input image to form a target image;
[0094] The recognition unit 230 is configured to input the target picture into the instance segmentation model, and output one or more coordinates of polygonal boxes enclosing the topic.
[0095] As an optional implementation manner, the creating unit 210 includes:
[0096] Constructing a subunit 211 for creating an initial model of instance segmentation based on deep learning, the initial model of instance segmentation includes a feature extraction network, a region candidate network, a regression network, and a segmen...
Embodiment 3
[0115] see Figure 4 , Figure 4 It is a schematic structural diagram of an electronic device disclosed in an embodiment of the present invention. Such as Figure 4 As shown, the electronic equipment may include:
[0116] a memory 310 storing executable program code;
[0117] a processor 320 coupled to the memory 310;
[0118] Wherein, the processor 320 invokes the executable program code stored in the memory 310 to execute some or all of the steps in the deep learning-based topic segmentation method in Embodiment 1.
[0119] The embodiment of the present invention discloses a computer-readable storage medium, which stores a computer program, wherein the computer program causes the computer to execute some or all of the steps in the method for topic segmentation based on deep learning in Embodiment 1.
[0120] The embodiment of the present invention also discloses a computer program product, wherein, when the computer program product is run on a computer, the computer is ...
PUM
Login to View More Abstract
Description
Claims
Application Information
Login to View More