Method and device for generating cover image, equipment and storage medium
A cover and image technology, which is applied in the fields of instruments, computing, electrical digital data processing, etc., can solve the problems of long-term refinement and low efficiency, and achieve the effect of improving the experience of watching videos, improving accuracy and efficiency
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
Embodiment approach 1
[0109] Implementation mode 1. Extraction mode.
[0110] Predicting first probabilities that at least one piece of text information is target description information based on the integrated text features, and then selecting at least one target description information from the at least one piece of text information based on the obtained first probabilities.
[0111] Specifically, the comprehensive text features are input into the encoder to obtain the first probability that at least one piece of text information is target description information, wherein the encoder can be ALBERT, BERT, Transformer-Encoder, etc. Sort the obtained first probabilities in order from large to small, and use the text information corresponding to the first probability greater than the screening threshold as the target description information, or the text information corresponding to the first probability ranked in the top N positions, As the target description information, N is a preset number.
[01...
Embodiment approach 2
[0116] Embodiment 2, generation method.
[0117] The integrated text features are encoded to obtain at least one second text feature corresponding to each of the text information. Decoding the second text feature corresponding to the at least one text information respectively, generating at least one candidate description information corresponding to the target video, and a second probability that each candidate description information is the target description information. Then, based on the obtained second probability, at least one target description information is selected from at least one candidate description information.
[0118] Specifically, an encoder is used to encode the comprehensive text feature to obtain at least one second text feature corresponding to each text information, wherein the encoder can be ALBERT, BERT, Transformer-Encoder, etc. A decoder is used to decode the second text feature corresponding to at least one text information respectively, and at l...
Embodiment approach 3
[0126] Embodiment 3: The extraction method is combined with the generation method.
[0127] In this embodiment of the present application, at least one target description information includes at least one first type of description information and at least one second type of description information, wherein the first type of description information is the target description information obtained by extraction, and the second description information is The target description information obtained by generating.
[0128] Based on the comprehensive text features, the process of predicting at least one target description information corresponding to the target video is as follows: Figure 7 shown, including the following steps:
[0129] Step S701, encoding the integrated text features to obtain at least one second text feature corresponding to each of the text information.
[0130] An encoder is used to encode the integrated text features to obtain at least one second text feature ...
PUM
Login to View More Abstract
Description
Claims
Application Information
Login to View More 


