孫旭菲 繆新穎 畢甜甜 王水濤 喻芳宇



摘要: 針對(duì)因光照、 拍攝角度及圖片質(zhì)量等因素導(dǎo)致的經(jīng)典深度學(xué)習(xí)算法難以有效提取人臉特征、 人物身份識(shí)別準(zhǔn)確率難以達(dá)到理想精度的問(wèn)題, 提出一種基于人臉強(qiáng)語(yǔ)義的年齡識(shí)別算法. 首先, 通過(guò)注意力矩陣增強(qiáng)人臉區(qū)域的特征權(quán)重, 達(dá)到提取特征區(qū)域的目的; 其次, 使用級(jí)聯(lián)雙向長(zhǎng)短期記憶(Bi-LSTM)網(wǎng)絡(luò)學(xué)習(xí)時(shí)序幀之間的特征依賴(lài)關(guān)系, 彌補(bǔ)部分特征缺失對(duì)識(shí)別精度的影響. 在人臉數(shù)據(jù)集IMDB-WIKI和數(shù)據(jù)集Adience上進(jìn)行測(cè)試, 該算法的年齡識(shí)別準(zhǔn)確率分別達(dá)到78.34%和77.89%. 實(shí)驗(yàn)結(jié)果表明, 相比于其他基于深度學(xué)習(xí)算法的方法, 該算法在基于圖片數(shù)據(jù)集的人物年齡識(shí)別任務(wù)上具有更高的準(zhǔn)確率.
關(guān)鍵詞: 年齡識(shí)別; 人臉識(shí)別; 深度學(xué)習(xí)算法; 注意力矩陣; 級(jí)聯(lián)Bi-LSTM
中圖分類(lèi)號(hào): TP391.41文獻(xiàn)標(biāo)志碼: A文章編號(hào): 1671-5489(2024)02-0347-10
SFSR-Age: An Age Recognition AlgorithmBased on Strong Facial Semantics
SUN Xufei1,2, MIAO Xinying1,2, BI Tiantian1,2, WANG Shuitao1,2, YU Fangyu1
(1. College of Information Engineering, Dalian Ocean University, Dalian 116023, Liaoning Province, China;
2. Key Laboratory of Liaoning Provincial Marine? Information Technology, Dalian 116023, Liaoning Province, China)
Abstract: Aiming at the problems that the classical deep learning algorithm was difficult to extract facial features effectively and the accuracy of character identification was difficult to reach the ideal accuracy due to factors such as illumination, shooting angle and image quality, we proposed an? age recognition algorithm based on strong facial semantics. Firstly, the feature weights of facial regions were enhanced by the attention matrix to achieve the purpose of extracting feature regions. Secondly, a cascaded bi-directional long short-term memory (Bi-LSTM) network was used to learn the feature dependency relationships between temporal frames and? compensate for the influence of missing features on recognition accuracy. When tested on IMDB-WIKI facial dataset and Adience dataset, the age recognition accuracy of the algorithm reached 78.34% and 77.89%, respectively. Experimental results show that compared with other methods based on deep learning algorithms, the proposed algorithm has higher accuracy in the task of person age recognition based on image datasets.
Keywords: age recognition; facial recognition; deep learning algorithm; attention matrix; cascaded Bi-LSTM
基于人臉的人物年齡識(shí)別算法在未成年人上網(wǎng)防沉迷、 網(wǎng)絡(luò)身份識(shí)別等領(lǐng)域應(yīng)用廣泛, 目前, 基于人臉年齡識(shí)別方法主要分為兩類(lèi): 基于圖像分析的方法和基于回歸分析的方法. 陳文兵等[1]提出了基于多模型融合的年齡識(shí)別模型CNN-SE-ELM, 該模型首先使用卷積神經(jīng)網(wǎng)絡(luò)(CNN)初步提取人臉特征, 然后使用SENet(squeeze-and-excitation networks)進(jìn)一步提取人臉深層特征, 最后使用EM-ELM(error minimized extreme learning machine)算法實(shí)現(xiàn)對(duì)人臉年齡的分類(lèi)及性別分類(lèi). 實(shí)驗(yàn)結(jié)果表明, 該算法能提取更有效的面部特征, 且分類(lèi)器能快速收斂. 陳濟(jì)楠等[2]提出了改進(jìn)的深度神經(jīng)網(wǎng)絡(luò)模型, 使用大卷積核以及跨連卷積實(shí)……