本文選題：人臉檢測 + 深度學(xué)習(xí)　； 參考：《華中科技大學(xué)》2016年碩士論文

【摘要】：人臉檢測是一個復(fù)雜的模式判別問題,其難點主要由成像角度不同所引起:如平面內(nèi)旋轉(zhuǎn)和平面外旋轉(zhuǎn),偏轉(zhuǎn)角度會直接影響判定人臉的準確度。當前基于深度學(xué)習(xí)卷積神經(jīng)網(wǎng)絡(luò)的檢測方法雖然有著很高的檢測率,但是在神經(jīng)網(wǎng)絡(luò)的輸出層對人臉的處理不夠精確,忽略了一張人臉對應(yīng)的多個檢測窗口之間的關(guān)聯(lián)關(guān)系,從而導(dǎo)致最終人臉框不夠精確。結(jié)合條件隨機場模型CRF對神經(jīng)網(wǎng)絡(luò)的輸出層進行調(diào)整,使得最終的人臉框更加精確。提出了一種基于卷積神經(jīng)網(wǎng)絡(luò)和條件隨機場模型的人臉檢測方法CRF-CNN,該方法提高了最終人臉框的精確度。方法首先對卷積神經(jīng)網(wǎng)絡(luò)進行訓(xùn)練,得到判定人臉和非人臉的分類器,對輸入圖像進行滑動窗口人臉檢測,得到包含人臉的窗口;然后標注同一張人臉對應(yīng)的所有檢測窗口,窗口對應(yīng)的置信分作為條件隨機場CRF的隨機變量,通過CRF模型計算窗口之間的關(guān)聯(lián)關(guān)系,根據(jù)關(guān)聯(lián)關(guān)系的緊密程度對窗口進行取舍;最后根據(jù)面積重疊的大小和橫向距離、縱向距離重疊的大小分別對同尺度和不同尺度的窗口進行合并,得到最終的人臉框。為了使得檢測率更高,該方法還對輸入圖片做了不同尺度的縮放處理,縮放程度的不同只會很小程度影響檢測時間,不會影響檢測的正確性,所以本方法對選用何種縮放算法及其參數(shù)并不敏感。實驗分別與卷積神經(jīng)網(wǎng)絡(luò)檢測方法DDFD、R-CNN和局部特征檢測方法DPM進行了比較。結(jié)果表明,CRF-CNN的準確率和召回率與DDFD相近,高于R-CNN和DPM。在面內(nèi)旋轉(zhuǎn)和面外旋轉(zhuǎn)的人臉檢測中,CRF-CNN得到的人臉框更加準確,尤其在面外旋轉(zhuǎn)的人臉檢測中,CRF-CNN置信分均值為0.99759,高出DDFD 0.00527。
[Abstract]:Face detection is a complex pattern discrimination problem, which is mainly caused by different imaging angles. For example, rotation in plane and rotation out of plane, deflection angle will directly affect the accuracy of face determination.Although the current detection method based on deep learning convolution neural network has a high detection rate, the processing of face in the output layer of neural network is not accurate enough, and the correlation relationship between multiple detection windows corresponding to a face is ignored.As a result, the final face frame is not accurate enough.Combined with conditional random field model (CRF), the output layer of neural network is adjusted to make the final face frame more accurate.A face detection method CRF-CNN based on convolution neural network and conditional random field model is proposed, which improves the accuracy of the final face frame.Methods firstly, the convolutional neural network was trained to obtain the classifier for judging face and non-face, and the sliding window face detection was carried out on the input image, and the window containing the face was obtained, and then all the detection windows corresponding to the same face were labeled.The confidence score of the window is regarded as the random variable of conditional random field CRF, the correlation relation between windows is calculated by CRF model, and the window is chosen according to the tightness of the correlation relation. Finally, according to the size of area overlap and the horizontal distance,The size of vertical distance overlaps is used to merge the windows of the same scale and different scales, and the final face frame is obtained.In order to make the detection rate higher, the method also makes different scale scaling of the input image. The different scaling degree will only affect the detection time to a very small extent, and will not affect the accuracy of the detection.Therefore, this method is not sensitive to the selection of scaling algorithm and its parameters.The experiments are compared with the convolutional neural network detection method DDFDR-CNN and the local feature detection method DPM.The results showed that the accuracy and recall rate of CRF-CNN were similar to those of DDFD and higher than those of R-CNN and DPM.In the in-plane rotation and out-of-plane rotation of the face detection, CRF-CNN is more accurate, especially in the out-of-plane rotation of the face detection, the average confidence score of CRF-CNN is 0.99759, which is higher than DDFD 0.00527.
【學(xué)位授予單位】：華中科技大學(xué)
【學(xué)位級別】：碩士
【學(xué)位授予年份】：2016
【分類號】：TP391.41;TP183

【參考文獻】

相關(guān)期刊論文 前2條

1 錢生;陳宗海;林名強;張陳斌;;基于條件隨機場和圖像分割的顯著性檢測[J];自動化學(xué)報;2015年04期

2 陳衛(wèi)中;潘曉平;倪宗瓚;;logistic回歸模型在ROC分析中的應(yīng)用[J];中國衛(wèi)生統(tǒng)計;2007年01期

，

本文編號：1743580