Object detection method and neural network training method, device and electronic equipment
A neural network and object detection technology, which is applied to biological neural network models, neural learning methods, character and pattern recognition, etc., can solve problems such as excessive calculation, detection technology has a large amount of calculation, and does not have scale invariance. The effect of ensuring the detection accuracy and reducing the amount of calculation
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
Embodiment 1
[0053] figure 1 is a flowchart of an object detection method according to Embodiment 1 of the present invention.
[0054] refer to figure 1 , in step S101, the data of the size range of the object is acquired from the image to be checked through the first neural network for detecting the size range of the object.
[0055] In the embodiment of the present invention, the first neural network can be any appropriate neural network that can realize feature extraction or target object detection, including but not limited to convolutional neural network, reinforcement learning neural network, generation network in confrontational neural network, etc. Wait. The setting of the specific structure in the neural network can be appropriately set by those skilled in the art according to actual needs, such as the number of convolutional layers, the size of the convolution kernel, the number of channels, etc., which are not limited in the embodiment of the present invention.
[0056] Where...
Embodiment 2
[0062] figure 2 is a flowchart of an object detection method according to Embodiment 2 of the present invention.
[0063] refer to figure 2 , in step S201, the data of the size range of the object is acquired from the image to be checked through the first neural network for detecting the size range of the object.
[0064] In an embodiment of the present invention, the data of the size range of the object in the image to be inspected may include a scale vector of the object in the image to be inspected, for example, a scale histogram vector in face detection. Each element of the scale vector respectively indicates the probability that the size of the object in the image to be checked falls within the size range corresponding to the element. In face detection, by using the image to be detected as an input of the convolutional layer of the first neural network, a scale response heat map of the image to be detected is obtained. Then, the scale response heat map is used as an ...
Embodiment 3
[0079] Figure 4 It is a flow chart of the neural network training method according to the third embodiment of the present invention.
[0080] refer to Figure 4 , in step S301, the detection data of the size range of the object in each of the sample images is acquired from a plurality of sample images containing object label information through the neural network to be trained.
[0081] During the training process of the neural network, by inputting multiple marked sample images into the neural network, the detection data of the size range of the objects in these sample images is obtained. Wherein, the neural network to be trained is the first neural network mentioned in the above embodiment.
[0082] Wherein, the neural network has multiple convolutional layers, and a global maximum pooling layer is set at the end of the last convolutional layer. By using the sample image as the input of the convolutional layer of the neural network, the scale response heat map of the sam...
PUM
Login to View More Abstract
Description
Claims
Application Information
Login to View More