【摘要】 深度神经网络是目前计算机机器学习领域的一个关键技术,可应用于图像处理。其中,多任务卷积神经网络(Multi-task Convolutional Neural Network,MTCNN)是一种基于卷积神经网络的多任务人脸检测框架,这里采用MTCNN人脸检测模型代替传统的卷积神经网络,在深度学习框架TensorFlow上进行人脸识别。首先,在数据预处理阶段利用灰度化方法将图像集转变为灰度图,降低图像通道。其次,基于MTCNN构建人脸检测模型,并利用Softmax函数进行分类识别。最后,实验过程中选择不同迭代次数进行准确性对比,在模型趋于稳定的情况下,得到较高的准确性。更多还原

【Abstract】 Deep neural network is a key technology in the field of computer machine learning, which can be applied to image processing. MTCNN(Multi-task Convolutional Neural Network) is a multi-task face detection framework based on convolutional neural networks, in which the MTCNN face detection model is used to replace the traditional convolutional neural network, and face recognition is performed on the deep learning framework TensorFlow. Firstly, in the data pre-processing phase, the gray set method is used to transform the image set into a grayscale image and reduce the image channel, then based on MTCNN, a face detection model is constructed, and the Softmax function is applied for classification and recognition, and finally, different numbers of iterations are selected for accuracy comparison during the experiment, and when the model becomes stable, higher accuracy can be obtained.更多还原