Face detection method based on cascaded connection convolutional neural network
A convolutional neural network and neural network technology, applied in biological neural network models, neural architectures, instruments, etc., can solve the problems of scarcity of positive sample data and inability to make full use of it, achieve good accuracy and speed, enhance image data features, The effect of reducing image noise
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
Embodiment Construction
[0021] The present invention divides the layers of the convolutional neural network into three layers, the first layer network is responsible for generating face candidate frames, the second layer network is responsible for further strengthening the face frames, and the third layer network screens the face frames and compares them with human faces. Further regression of the face frame.
[0022] A kind of face detection method based on cascade convolutional neural network of the present invention comprises the following steps:
[0023] In the first step, after image preprocessing, a 12*12 three-channel image is input, and a face candidate frame is quickly generated through a fully convolutional neural network algorithm to provide discriminative data for the second-level input. The neural network at this level finally divides the input area into three categories—face area, partial face area, and non-face area, and predicts the relative position of the face frame of the face area...
PUM
Abstract
Description
Claims
Application Information
- R&D Engineer
- R&D Manager
- IP Professional
- Industry Leading Data Capabilities
- Powerful AI technology
- Patent DNA Extraction
Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.
© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com