Target identification method and device
A target recognition and equipment technology, applied in the field of target recognition, can solve the problems of network loss of regularity, unfavorable parallelism, etc., achieve the effect of maintaining the same network performance, increasing the degree of tailoring, and speeding up the training speed
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
Embodiment 1
[0043] This embodiment provides a method for target recognition, such as figure 1 As shown, it specifically includes the following steps:
[0044] Step 101, acquiring image data for target recognition;
[0045] Step 102, input the image data into the neural network model for target recognition, and obtain the recognition result of the target recognition, wherein, the neural network model is obtained by inputting the training samples into the preset neural network model for training, and the preset neural network model is In the process of training the neural network model, it is determined whether the preset neural network model includes a batch normalization BN layer; if so, the number of layers of the BN layer in the preset neural network model is cut, otherwise, all The weights of the network layers of the preset neural network model are clipped.
[0046] In the above method, by combining the direct and indirect structured sparsity methods with the preset neural network m...
specific Embodiment approach
[0051] If it is determined that the current preset neural network model includes a BN layer, the number of layers of the BN layer of the current preset neural network model is trimmed. The specific implementation is as follows:
[0052] As an optional implementation manner, the number of layers of the BN layer of the preset neural network model is trimmed by adding a second penalty item to the BN layer of the preset neural network model.
[0053] The second penalty item includes a second adjustment coefficient for adjusting the output value of the BN layer in the preset neural network model and a range of the output value of the cropped BN layer;
[0054] For the BN layer in the preset neural network model, adjust the output value of the BN layer in the preset neural network model according to the second adjustment coefficient described in the second penalty item;
[0055] By dynamically adjusting the second adjustment coefficient in the second penalty item, the output value ...
Embodiment 2
[0073] Based on the same inventive concept, this embodiment provides a target recognition device, such as image 3 As shown, the device includes a processor 301 and a memory 302, wherein the memory 302 stores an executable program, and the processor 301 implements the following process when the above-mentioned executable program is executed:
[0074] Obtain image data for object recognition;
[0075] Input the image data into the neural network model for target recognition, and obtain the recognition result of the target recognition, wherein, the neural network model is obtained by using the training samples input into the preset neural network model, and the preset neural network is In the process of model training, it is determined whether the preset neural network model includes a batch normalization BN layer; if so, the number of layers of the BN layer in the preset neural network model is cut, otherwise, the preset The weights of the network layers of the neural network ...
PUM
Login to View More Abstract
Description
Claims
Application Information
Login to View More